MXPA97002942A - Method and apparatus for detecting and predicting motion of mobile terminals - Google Patents

Abstract

Methods and apparatus for detecting and predicting movement patterns of mobile radio transceivers, such as mobile cellular telephones. One method of predicting a next location of a mobile terminal based includes the step of comparing a current sequence, that includes the current location of the mobile terminal, and a plurality of previous locations of the mobile terminal to each of a plurality of stored sequences that each include previous locations of the mobile terminal. The method also includes the steps of selecting one of the stored sequences based on at least one quantitative measure of a degree of matching between the current sequence and each stored sequence, and predicting the next location of the mobile terminal based on the selected one of the stored sequences. Methods and apparatus for determining regular patterns in movements of a mobile terminal are also described;as well as is a communication network with several servers, wherein a mobile terminal has a device for communicating with the nearestáserver. The device accesses application and data files stored in the servers. Also described is a mobile distributed system platform having a device for controlling the distributed file system, but also a device for predicting a next location of a mobile terminal, wherein this device distributes location related information among the servers, based on a next location predicted being predicted.

Description

METHOD AND APPARATUS FOR DETECTING AND PREDICTING THE MOVEMENT OF MOBILE TERMINALS BACKGROUND OF THE INVENTION The present invention of the applicants refers to the systems and methods for predicting the movements of mobile radio transceivers, such as mobile cellular telephones. A fundamental shift is taking place, it is the movement to communication and mobile computing together in a synergistic activity - mobile computing - a new dimension in mobile communication networks. Future mobile communication systems will support e-mail services, access to database and files, and multimedia services in addition to well-known voice communication. Due to non-uniform traffic patterns of mobile terminals in different geographical areas, stratified or hierarchical cellular architecture is a promising solution for mobile communication systems in the future. In general, a hierarchical cell structure consists of layers of orthogonal cells of different sizes and cell types that cover a common geographic area. A hierarchical cellular structure usually consists of macrocells that overlap microcells and picocells, which are distinguished by their extension n space, among other features. For example, u: a picocell is an area that has a certain number of groups of channels or an identical code, and the nominal cell radio of a picocell is usually less than 200 meters. In a stratiform communication system, the width of the communication of a pico cell, which is common in 'interiors, can be up to 2-10 megabits per second (Mb / s), while the width of Communication band of a macrocell is in the order of 10-100 kilobi s per second (Kb / s). However, many challenging aspects can be solved to support continuous, transparent communication for the user for both voice and data communication across the borders of the different cell layers. The connectivity and configuration of these wireless networks is highly dynamic because mobile terminals can change their positions and their radio environments can change at any time. Also, due to the large difference in bandwidth between the cell layers, the communication networks that are currently available are not transparent to mobile data users. Conventional techniques for entering reserve memory and data pre-search are not enough in this environment. The introduction of data in back-up memory and pre-search are commonly used to improve system performance in large-scale distributed computing systems. Statistics indicate that even small reserve memories provide substantial benefits. Backup or cache memory can not only reduce latency or wait time but can also greatly reduce the volume of traffic on the network, as well as the number of server uses in a client-server system. The pre-search is complementary to the reservation memory, and the successful pre-search of the data for the local reservation memory will increase the speed of certain of the memory, that is, the frequency with which the requested damages are found in the memory. reserve memory. The idea of using standby memory to reduce waiting time and traffic on the network is based on the properties of the temporary location of the access patterns to the data of the computer programs. This fact is described in D. Lilja, "Cache Coherence in Large-Scale Shared Memory Multiprocessors: Issues and Comparisons", ACM Computing Surveys, vol. 25, no. 3. Pp. 303-338 (Sept. 1993). Temporary location means that it is very likely that the data newly accessed by a program will be accessed again in the immediate future. In this way, a local copy or cache of the remote data that has been recently accessed is retained, so that repeated access to the same data can be handled locally without additional traffic in the network. With a well-managed cache, you can access large amounts of soaking data with almost the same efficiency as local data. On the other hand, the existence of multiple copies of shared data introduces the problem of maintaining the consistency of the copies described in the Lilja publication described in the above. When the data entered in the backup memory is modified, changes must be made in all local and remote copies of this data. The various consistency schemes of the backup memory can be used to keep the copies consistent according to the data sharing semantics used, but maintaining the consistency of the backup memory in distributed systems is a very complicated problem that includes the tactical balance between consistency, transparency and traffic volume in the network.

Pre-search is another useful technique in distributed computing systems that uses knowledge about the expected needs for remote data. The remote data is pre-searched for a client, that is, it is searched before the remote data is actually requested. The knowledge of the expected needs can be determined from the client's previous behavior. For example, operational file games are often the ones that are searched in distributed file systems, and operational games are determined based on knowledge of the pattern of access to client files. Another common pre-search method uses spatial locality, which refers to the high probability that a customer needs the data stored in the neighboring addresses or pages of data already needed. At present, networks are not efficient in the wireless access of data since they do not support the mobility of data and service. Although users and terminals are mobile, their data is still statically configured in the system. Traditionally, the management of personal / terminal mobility included passive functions to keep track of the locations of the users or terminals and to maintain the connection with the terminals that belonged to the system. Also, due to the large differences in the bandwidth of data communication between cellular layers, the communication system is not transparent for mobile data users. In other words, the different cell layers have large differences in performance for mobile data users. Conventional techniques for managing the data entry in the backup memory and the prefetch are mainly designed for fixed data communication networks, and are inefficient in a communication medium, such as a cellular radiotelephone system, in which Communication channels are not predictable and are highly variable with time and location. The case is, how to improve the performance of the network in a hierarchical cellular system, and especially, how to make the use and administration of the network more intelligent balancing the traffic in the network and the allocation of dynamic channels, how to make smart the management the introduction of data in the backup memory and the pre-search of data in the mobile environment, and how to provide wireless access to data efficiently.

Therefore, techniques that can predict the movement (itinerary) of a user with a mobile terminal are necessary to support hands-on, or hands-off connection between the different cellular layers with high integrity.

COMPENDIUM OF THE INVENTION These problems are solved by the invention of the applicants, which provides the methods and apparatus for predicting the movements of mobile radio transceivers. In this way, the invention of the applicants achieves the important objectives of improving the functioning of a hierarchical communication network, improving the use of the network and the balanced handling of traffic in the network and the allocation of dynamic channels and the improvement of the introduction of the data in the reserve memory and the pre-search of the data in a cellular mobile communication network. In one aspect of the applicants' invention, a method for predicting a next location of a mobile terminal based on the stored prior locations of the mobile terminal includes the step of comparing a current sequence that includes the current location of the mobile terminal. and a plurality of previous locations of the mobile terminal for each of a plurality of stored sequences each including prior locations of the mobile terminal. The method also includes the steps of selecting one of the stored sequences based on at least one quantitative measure of a degree of coincidence between the current sequence and each of the stored sequences., and the prediction of the next location of the mobile terminal based on one of the stored sequences. In another aspect of the applicants' invention, a method for predicting the movements of a mobile terminal includes the steps of: (a) comparing a current sequence that includes the current location of the mobile terminal and a plurality of prior locations of the mobile terminal with each of a plurality of stored sequences, each of which includes the previous locations of the mobile terminal; (b) determining at least one quantitative measure of a degree of coincidence between the current sequence and each of the stored sequences; and (c) if at least one quantitative measure exceeds a predetermined value, the use of the locations of the respective stored sequence as predictions of the movements of the mobile terminal. In another aspect of the applicants' invention, a method for determining normal patterns in the movements of a mobile terminal includes the step of comparing a current location of the mobile terminal for each of a plurality of previous locations stored in a queue. When waiting for a plurality of previous locations, the previous locations are stored in the waiting queue in the order of appearance of first entries-first outputs. The method also includes marking a sequence of locations comprising: the current state, the previous location that matches the current location and the previous locations that were presented after the previous location that matches the current location, if the current location matches one of the plurality of previous locations stored in the waiting queue. The method also includes comparing the tagged sequence with each of a plurality of stored sequences from the locations and determining at least one quantitative measure of a degree of coincidence between the tagged sequence and each of the stored sequences, and increasing a priority parameter of the respective stored sequence by a predetermined amount if at least one of the quantitative measurements exceeds a predetermined value. This method can also include the storage stages, in the waiting queue, of the current location of the mobile terminal in the order of appearance of first inputs-first outputs, determining whether the current location is at least a stationary state. or a boundary state, and carrying out the other steps if the current location is at least a stationary state or a boundary state; (d) the marking of a sequence of locations consisting of: the current location, one of the most recent previous stationary states and the most recent previous boundary state, and the previous locations that occurred between the most recent previous steady state and the most recent previous boundary state; (e) comparing the marked sequence with each of a plurality of stored sequences of the locations and determining at least one quantitative measure of a degree of coincidence between the marked sequence and each of the stored sequences; and (d) if at least one quantitative measure exceeds a predetermined value, the increment of a priority parameter of the respective stored sequence by a predetermined amount. This method can also include the storage stage of the current location in a waiting queue of a plurality of previous locations, the locations are stored in the waiting queue in the order of appearance of first entries-first outputs, and carrying out the other stages if the current location is at least a steady state or a boundary state.

In these methods, the plurality of stored sequences may include movement cycles and movement tracks, and one of the stored sequences may be selected based on a ratio of the number of locations in the current or marked sequence that are the same as the locations in a stored sequence and a total number of locations in the current or marked sequence, Likewise, one of the stored sequences may be selected based on a quantitative measure of the degree to which a current or marked sequence duration matches the duration of each one of the stored sequences and in a quantitative measure of the degree to which the frequency of the current or marked sequence coincides with the frequency of each of the stored sequences. In other aspects of the applicants' invention, an apparatus is provided for predicting the next location of a mobile terminal based on the previous locations of the mobile terminal, and the apparatus for determining the normal patterns of the movements of a mobile terminal. . For example, an apparatus for predicting the next location of a mobile terminal based on the previous locations of the mobile terminal includes a current one coinciding with the duration of each of the stored sequences, and a device for generating a third quantitative measure of the degree in which a frequency of the current sequence coincides with a frequency of each of the stored sequences. One of the sequences of the stored sequences can be selected based on the ratio and the second quantitative measure, or, in the relation, the second quantitative measure and the third quantitative measure. In still another aspect of the invention of the applicants, a communication network consists of a plurality of servers, the servers are placed in respective geographical areas and organized in a distributed file system; a mobile terminal has a device for communicating with the server closest to the mobile terminal, wherein the communication device has access to the application files and the data files stored on the servers; and a mobile distributed system platform having a device for controlling the distributed file system of the servers and a device for predicting a next location of a mobile terminal, wherein the control device distributes the location sensitive information between the servers with base in a next location predicted by the predictive device.

BRIEF DESCRIPTION OF THE DRAWINGS The characteristics and advantages of the invention of the applicants will be understood by reading this description together with the drawings, in which: Figure 1 is an example of the hierarchical or multilayer cellular system; Figure 2 is a block diagram of an example of the cellular mobile radio system in which an example of the base station and the mobile station is included; Figure 3 is an illustration of the movement of a mobile in a hierarchical communication system; Figure 4 illustrates the MMP apparatus of the applicants, which serves to predict the movement of the mobile; Figure 5 illustrates the way in which the user moves through the different states that are grouped in cycles of movement; Figure 6 illustrates how the states are grouped into movement tracks; Figure 7 illustrates the operation of an itinerary pattern detector and the data structures that are generated in an itinerary pattern database; Fig. 8 is a flowchart of a method for detecting the movement cycle according to the invention of the applicants; Figure 9 is a flow diagram of a method for detecting the movement tracks according to the present invention of the applicants; Figure 10 illustrates the operation of a motion predictor; Figure 11 illustrates the structure, similar to a tree, of a method for predicting movement; Figure 12 illustrates an example of how the user moves, through the different states, as a function of time; Figures 13a, 13b illustrate an example of the constitutional constraints that are used in the matching processes; Fig. 14 in a flow chart of a movement predictor of a mobile according to the present invention of the applicants; Figure 15 shows an example of the standardized results of a simulation of the movement predictor of a mobile of the applicants; and Figure 16 illustrates a mobile floating agent and a platform agent of the mobile distributed system employing the motion predictor of a mobile of the applicants.

DETAILED DESCRIPTION Most people, including most users of mobile terminals, have normal patterns of movement that continue more or less daily during the week. The invention of the applicants makes use of this regularity in the movements of each one to predict the next location of a person. For example, if a user of a mobile is almost out of the area covered by a picocell (or microcell), the mobile (or network) can predict this change of location and inform the network to pre-search the data, if necessary . Using the invention of the applicants, a mobile terminal or the communication network can predict the itinerary of the mobile and take the appropriate actions before the mobile arrives at a new location. These predictions can also be used for the allocation of dynamic channels, the location of the mobile terminal and the calls hands out from one channel to another, either intra-cellular or inter-cellular, inter-layer or intra-layer. Predictions can be entered into a localization algorithm, which generates a list of the candidate communication channels for hands over or assignment of a connection. When used in this application, the term "mobile terminal" shall be understood to include mobile phones, laptops, mobile texts, personal digital assistants, and the like. Figure 1 is an example of a hierarchical or multilayer cellular system. An umbrella macrocell 10 which is represented by a hexagonal shape constitutes a superimposed cellular structure. Each umbrella cell can contain a fundamental microcellular structure. The umbrella cell 10 includes the microcell 20 which is represented by the area enclosed within the dotted line and the microcell 30 is represented by the area enclosed within the dotted line corresponding to the areas along the streets of the city, and the picocells 40, 50 and 60, which cover the individual floors of a building. The intersection of the two streets of the city that are covered by the microcells 20 and 30 can be an area of dense traffic concentration, and thus could represent a high point. Figure 2 represents a block diagram of a mobile cellular radio system including an example of a base station 110 and a mobile station 120. The base station includes the control and processing unit 130 which is connected to a mobile switching center ( MSC) 140 which in turn is connected to the public switched telephone network (PSTN) (not shown). The general aspects of these cellular radiotelephone systems are well known in the art, as described, for example, in U.S. Patent No. 5,175,867 to Wejke et al., Entitled "Neighbor-Assisted Handoff in a Cellular Communication System", and U.S. Patent Application No. 07 / 967,027 entitled "Multi-mode Signal Processing" filed on October 27, 1992, both are incorporated herein by reference. The base station 110 handles a plurality of voice channels through a voice channel transceiver 150, which is controlled by the control and processing unit 130. Likewise, each base station includes a control channel transceiver 160, which may be able to handle more than one control channel. The control channel transceiver 160 is controlled by the control and processing unit 130. The control channel transceiver 160 transmits the control information on the control channel of the base station or cell to the blocked mobiles for this control channel. It should be understood that the transceivers 150 and 160 can be installed as a single device, as can the voice and control transceiver 170, for use with digital control and traffic channels that share the same radio carrier frequency.

The mobile station 120 receives the transmission of the information in a control channel in its voice transceiver and voice channel 170. After, the processing unit 180 evaluates the information received in the control channel, which includes the characteristics of the cells that are candidates for blocking the mobile station and determines in which cell the mobile should be blocked. Advantageously, the information received from the control channel not only includes the absolute information related to the cell with which it is associated, but also contains relative information related to other cells next to the cell with which the control channel is associated, as it was described in U.S. Patent No. 5,353,332 to Raith et al., entitled "Method and Apparatus for Communication Control in a Radiotelephone System", which is incorporated herein by reference. An example of a movement pattern of the mobile terminal is that shown in Figure 3. A mobile user A moves through an area covered by a hierarchical cellular architecture, which includes a system of picocelalas having a width of 2 Mb / s communication band a macrocell system, such as a GSM system, with a bandwidth of 9.6 Kb / s for data transmission. User A entered one of the areas covered by the pico cell through door D and moved into the area covered by the pico cell (which may be inside a building) for a period of time, user A entered room L, moved into conference room C and then left the first area covered by the pico cell by door D, entering the area covered by the macro cell, as shown in the figure. User A moved through the area covered by the macroceluia to another area covered with pico cells, entering through another door B and moving within this area. According to one aspect of the invention of the applicants, the user A's itinerary must be recorded as in a Mobile Motion Predictor (MMP) in the mobile terminal of the user a or in the network. When user A moves with a certain speed towards the PLUS point from room C or room L, the MMP must indicate a high probability that user A leaves the area covered by the picocell, of high bandwidth . The MMP must inform other systems and applications that take appropriate action, such as the allocation of dynamic channels and the pre-search of data, if necessary, before the user leaves. As illustrated in Figure 4, the MMP of the applicants consists of an IPD route pattern detector, an IPB route pattern database, and an MP movement predictor. The IPD is used to detect the normal route patterns (IP) in the movements of a user between locations or states and to store the IPs in an itinerary patterns database (IPB). In general, the IPB also includes predetermined information regarding the constitution or physical structure of the communication system, as will be described in more detail later. The MP uses the information of the route patterns stored in the IPB to predict the next location or the user's status. The MP also compares its predictions with the user's next real states and updates the IPs stored in the IPB. The input data provided by the MMP are the IA, or states, in which it is located in mobile, and currently it is believed that the system must be continuously verified for a new state with a predetermined period, for example, every 1- 5 sec. It should be noted that the IAs identify the locations of the mobile, that is, the cells where the mobile has been and has been located. In this way, AIs can have any suitable form, such as code numbers in a multiple access communication system to the division of codes (Code Division Multiple Access, CDMA), or cell locations in an access communication system. Multiple to the division of time (Time Division Multiple Access, TDMA) such as the GSM system used in Europe and the AMPS system used in the United States. The itinerary patterns or AI sequences are stored in the IPB to which it is accessed by means of the movement predictor MP. For the analysis of the correlation of the CM or MT, three kinds of correspondence schemes are used. First-class correspondence or state correspondence indicates the degree to which a sequence of states matches other sequences of states with a similar length, this is quantified by a first-class matching index as described in the following. The second-class correspondence or velocity or time correspondence indicates the degree to which the duration of a sequence of states coincides with the duration of another sequence of states with a similar length. This is quantified by the second class matching index as described below. The third-class correspondence or frequency correspondence indicates the degree to which the frequency of a sequence of states coincides with the frequency of another sequence of states with similar length; it is quantified by the third class correspondence index described below.

Before describing the invention of the applicants in more detail it will be useful to point out the following concepts and abbreviations: MC limit (boundary MC, BMC): movement cycle (MC) in which at least one state is a limit state; a BMC has a higher priority than an MC, with a limitation of priority parameter ß. MT lim it (MT boundary, BMT): a movement track (MT) in which at least one state is a limit state; a BMT has a higher priority than an MT, with a limiting priority parameter ß. Limit state (boundary sta te, BS): a state at the boundary of the cell layer. FIFO (FIFO): first inputs-first outputs. State forked (forked sta te): a united state for which the following states are in distinguishable movement cycles. Identity area (identi ty area, IA): a cell or group of cells that send (transmit) identified information to an area covered by the cell or group of cells. Base of the i tinerari sheet or (i tínerary pa t tern base, IPB) an information database that includes a maximum number MORE of itinerary patterns.

United State (J oint sta te, JS): state that is included in at least two distinguishable movement cycles. LRU: the newly used. cycle of movement (movement cycle): a cycle that has n (where n> 1) sequential states that includes at least a steady state. Movement trail (movement track, MT): track that begins and ends with a stationary state or a limit state. Pointer State (PS pointer): A state in a state wait queue that contains a pointer pointing to an MC or MT stored in an IPB. p: priority parameter, indicates the priority of an MC or MT. Estaao (s ta te, S): location of a user, that is, an area of identity IA, in a movement pattern (or movement graph), where Sk, t indicates the state k in time as well (current time ). State wait queue (sta te queue, SQ): a wait queue of states that were stored in order of appearance. state is taci onari or (sta ti onary sta, SS): a state (IA) in which the mobile terminal has remained for at least a period of time t.

Tmc: period of an MC given by tn - ti, the time interval between the first and last state in the MC. Transition state (transition, TS): a state in which a mobile terminal has remained for less than a period of time t. tB: time criterion to identify a BS. rs; cri terio of time to identify an SS.

According to the invention of the applicants, the IPD is based on two fundamental procedures: a model of movement cycles (MC) and a model of movement tracks (MT). The MC model is responsible for the long-term normal user movements, which are assumed to take the form of loops or closed states cycles. The MT model addresses routine movements, which are assumed to take the form of linear clues of states. The MC model is based on the assumption that when a user moves from a place, the user will finally return ,, In this way, the movements of a mobile terminal user are modeled as different patterns similar to circles, the examples of which are they are represented in figure 5. The states are represented by circles identified with numbers 1-27, 19-35 that indicate the areas of identity IA that correspond to the states. From the figure it can be seen that an MC is a closed loop or "cycle" of states having a duration Tmc that includes at least two states and at least one steady state. The MMP of the applicants determines that a state is a steady state when the following criterion is applied: if the signal IA (input state) provided to the MMP has not changed during the predetermined time period ts (eg, ts> 5 minutes ), then the Skft state is a stationary state. Looking at it in another way, each cycle of movement is a sequence of states, for example, [1, 16, 17, 18, 21, 20, 19, 18, 1]. It should be underd that when considering a movement cycle, one must move around the circle in a known direction because the order of the states can change for different directions. Also, associated with each MC is a priority parameter p of the LRU that indicates the priority of the MC with respect to e. another MC in the IPB, a frequency parameter F, which indicates the frequency of the sequence of states (see Figure 12), and a limit priority parameter ß, all of which are initialized to 0 for each new MC. Ur is detected "new" MC comparing an MC that enters with each one of the MC already ed in the IPB. If the μ index of first-class correspondence of the new CM, which will be described in greater detail later, matches the index μ of one of the ed MCs, the priority factor p of the ed MC, the priority factor p of the MC ed increases 1. Otherwise, the new MC is ed in the IPB with the initial values of p = F = ß = 0. If one or more states of an MC is a limit state, this MC is called MC limit and has its limit priority parameter ß increased by 1. The MMP of the applicants determines that an input status is a limit state by applying one of the following criteria: either (1) if no input signal IA has been received for a period of time default time tb (for example, tb = 5 minutes), then the state Sk, t -. b is a limit state; or (2) if no input signal IA has been received after a predetermined time period tb (eg, tb = 5 minutes) and a new signal IA (state Sv +?,) has been received, then the state Sk + i, also is a limit state. Using the MT model, the IPB generates movement tracks, which are itineraries that each one starts and ends either with a stationary state or a limit state. Since closed loops are not needed, the MT model is a less limited version of the MC model.

Figure 6 shows six examples of the MTs that are derived from the MCs shown in Figure 5, and it should be noted that each MC includes at least one MT. As with movement cycles, it should be understood that when considering a movement track, one must move along the track in a known direction because the order of the states can change for different directions. Also, associated with each MT are the priority parameter p of the LRU, which indicates the priority of the MT with respect to other MTs, the frequency parameter F and the limit priority parameter ß, all of which are initialized to 0 for every new MT. A "new" MT is detected by comparing an MT that enters with each type of MT previously stored in the IPB. If the corresponding first class μ index of the new MT, which will be described in greater detail later, matches the μ index of one of the stored MTs, the priority parameter p of the stored MT will be increased by 1. Otherwise, the new MT is stored in the IPB with the initial values of p = F = ß = 0. If one or more states in an MT is a limit state, this MT is called a limit MT and has its limit parameter ß increased in 1.

The operation of the IPB and the data structures generated in the IPB will be better understood in connection with Figure 7, which illustrates an MCI movement cycle comprising three movement tracks MT1, MT2, MT3 stored in the IPB in the order of the most recently used (LRU). Also in Figure 7 is shown a waiting queue of SQ states that contains the most recent N states provided to the MMP, stored in the order first inputs-first outputs (PEPS) (read from left to right in the figure). The arrows indicate how the IPB transforms the sequence of states into a queue of states in the MTs stored in the IPB (which does not correspond to figures 5 and 6). The signal "C: 1/8, Wall, Street, Highway, etc." it refers to the examples of the constitutional restrictive states that arise from the physical construction of the communication system, which will be described in more detail later. It should be understood that IPB and SQ can be instrumented in a wide variety of conventional electronic memory circuits.

MC Detection Method (MCD) When the itinerary patterns are generated based on the MC model from the states that are in the state waiting queue, the IPB carries out a method of detecting movement cycles (DCM) consisting of the following steps, which are also illustrated in the flowchart of Figure 8. The DCM and the methods according to the invention of the applicants are described in terms of the C language pseudocode, by which methods can be easily implemented in the hardware and software in any of the mobile stations, mobile stations and mobile switching centers of a cellular radiotelephone communication system. Let N be the maximum number of states in the SQ state waiting queue; be MC- the MC jth; and be the other terms as defined in the above. Maintain a waiting queue for the k states (where 1 <k <N) in an order PEPS, and a number j of MC (where 1 <j <M) in a replacement order LRU in the IPB according to the following steps. BEGIN 1) IF Sk / t is a steady state or a limit state, IF FOR j = k = L to 1, if any Sift == Sk, t (for any t, and for k-i >; L) y If, t is an SS or a BS, mark the sequence of states [S?, T-tu Sm, t-t2, ..., Sk (t] as a new MC; ENDIF; ELSE GOTO END; ENDIF; 2) IF any state in the new MC is a limit state, mark the new MC as "limit priority" with a limit priority parameter ß - ß + 1; ENDIF; 3) Compare each new MC with each stored MC, that is, each MC already stored in the IPB, IF μ = al (correspondence), then increase 1 to the priority parameter p of the stored MC and calculate the frequency parameter F of the MC stored; ELSE IF a2 < μ < al (partial correspondence), then mark the last state that coincided as "forked states" in both MCs; ENDIF; Store the new MC in the IPB in the replacement order LRU; ENDIF; Remove the sequence [S., t-t-f Si + ?, t-t2, ..., Sk, t] from the state waiting queue; END. In the aforementioned, a2 and a are numbers that can be 0 < a2 < to < 1; SS are stationary states; BS are limit states; L = 1, 2, 3, ... is the length (in number of states of the shorter MCs stored in the IPB, and μ is the first class matching index.) Parameter a is a level of confidence that is It establishes according to the accuracy or level of confidence desired, generally, when set to 0.95, 0.975, 0.99, or similar.The parameter a2 is a partial correspondence factor that represents the degree of correspondence between two sequences of states, so that a2 = 0.3, 0.4, 0.5, ... corresponds to 70%, 60%, 50- ^, ..., of the states that coincide; it is believed that a2 should be set to at least 0.5 due to the interesting results that seem to be currently occurring, only when at least half of the states coincide in two sequences. The μ index of first-class correspondence is an indicator of the degree to which a first sequence of states coincides with a second sequence of states having a similar length (state correspondence). The μ index is determined by the following expression: «« enters formula, p. 18 »» where €: m2 is the number of states that match the sequence, and n2 is the total number of states in each sequence. Motion track detection method (MTD) When you generate route patterns based on the MT model, the IPB carries out a motion track detection (MTD) method consisting of the following stages, which are also illustrated in the flow diagram of Figure 9. Let MTj be the jth MT; let M be the maximum number of MT in the IPB; and be the other parameters as defined in the above. Keep a queue waiting for the k states (where 1 <k <N) in an order PEPS, and a number j of MT (where 1 < j < M) in a replacement order LRU in the IPB. BEGIN IF Sk, r is a steady state or a limit state, IF FOR i = k - L to 1, if Slft sa SS or BS (for any t, and for k-1> L), mark the sequence [S- .t-ti / Si.-ir-t ?, ..., Skft] as a new MT; ENDIF; ELSE GOTO END; ENDIF; 2) IF any state in the new MT is a limit state, mark the new MT as "limit priority" with a limit priority parameter ß = ß + 1; ENDIF; 3) Compare each new MT with each MT already stored, IF μ > al (correspondence), then increase 1 to the priority parameter p of the stored MT and calculate the frequency parameter F of the stored MC and calculate the frequency parameter F of the stored MT; ELSE IF a2 = μ < al (partial correspondence), then mark the last state that coincided as "forked states" in both MTs; ENDIF; Store the new MT in the IPB in the replacement order LRU; ENDIF; replace the last MT stored with a PS in the state queue; END. In the aforementioned, μ is the index of first class correspondence; a2 and a are numbers, such as 0 < a2 < to < 1; PS is a pointer state; L = 1, 2, 3, ..., is the minimum length (in number of states) ie an MT. It should be noted that the use of a pointer state instead of another state is advantageous since it avoids duplicity. Motion Predictor The MP movement flag included in the MMP of the applicants generates predictions of the next states of motion cycles or tracks of movement using regression analysis and correlation of the current movement itinerary with the IPs stored in the IPB. In general, the PDout output of the MP is the future state or a sequence of future states. Figure 10 illustrates the operation of the MP, which includes e .. means to compare the input states provided to the MMP to predict the states generated by the MMP and the means to match the input states with the IPs stored in the 1PB and for general predictions. If the comparator indicates that a prediction is correct, that is, that the current input state matches the predicted state, the prediction is proportioned as the output of the MMP. If the comparator indicates that the current input state does not match the prediction or when the MMP is initialized, a motion prediction process is carried out to generate the next prediction. When an input state does not match the corresponding predicted state, the sequence of the input states beginning with the steady state or the most recent limit state are compared with the MP for each of the MC and MT stored in the IPB. This correspondence process determines the stored route pattern that best corresponded, which is the output of the movement predictor. The motion prediction method, which consists of the following stages, is advantageously structured as a tree, as illustrated in Figure 11. Method for motion prediction (MPM) Let Sμ, t the state k in time as well, and be N >; 0, tj > 0, and let M be the maximum number of MT and MC in the IPB. Keep a waiting queue of the k states (where 1 <k = N) in a FIFO order and keep a number j of MT and MC (where 1 <j <M) in the IPB in a replacement order LRU . Also, suppose that PD, lt - [OJ, or, PDout? [0], and Sk, does not correspond to the first state of PD0Ut • EIGIN 1) FOR each new state that enters Sk, t compare the new sequence [Sk-nff-rn, Sk-n +? T-tn +? , ..., Sk / t] with each MC and MT stored in the IPB (where Sk-n, t-tn is the last SS or BS, and N> 0); ) Use first class correspondence (μ): IF only a stored MC or MT has fulfilled the requirement of correspondence μ with the new sequence (μ = al), then PD, ut = [Siui. / S +2 + t2, ..., S .m, tttm] of the corresponding MC or MT, (M> 0); GOTO END; ELSEIF the MC or MT not stored have a μ that matches the μ of the new sequence, then PDOÜ, - C "uc; GOTOEND; ELSE use second-class correspondence (?); ENDIF; ) use second-class correspondence (?); FOR all MC or MT stored that has μ that matches the new sequence (μ = al) (united states), I only an MC or MT stored has fulfilled the correspondence requirement? with the new sequence (? = a3), then PD "r = [SkU, MM., Skí ,, t.t2, ..., Sk + m, t + mJ of the corresponding MC or MT, (m > 0); GOTOEND; ELSEIF MC or MT not stored has a? What does it match? of the new sequence, find one that has a restrictive state that better matches the μ index and that matches the index better? of the new sequence; GOTOEND; ELSE use third-class F correspondence; ENDIF; use third-class F correspondence; FOR all stored MCs or MTs that match μ and? (U.S); if only an MC or MT stored has fulfilled the correspondence requirement F, PDout - [Sk.?, t + ti / SM2ft.t? / .... Sk + m, t + rm] of the MC or MT that matches (m> 0); GOTOEND; ELSEIF no stored MC or MT has an F that corresponds to the F of the new sequence (F = a4), look for one that has a restrictive state and that its indices μ,? and F coincide better with the new sequence; GOTOEND; ELSE (more than one MC or MT stored has an F that matches the F of the new sequence (F = a4), look for a restrictive state that has the highest p + ß, ENDIF, END. is the first class correspondence index,? is the second class correspondence index and F is the third class correspondence index, and the other terms are as described in the above. ? = a3 when there is correspondence with?, and have F < a4 when there is correspondence with F, where a3 and a4 are the confidence levels associated with the second and third and third class correspondences, respectively, which coincide with The velocities or frequencies of two sequences of states Since the speed or frequency of a mobile user is usually more highly variable, the values of the parameters a3, a4 do not need to be as restrictive as the values of the parameters meters to, a2. In this way, the values of a3 and a4 can be 0.1, 0.05, 0.025, 0.005, ..., depending on the requirements of accuracy or level of confidence desired; In general, a3 and a4 can be set to 0.05 for a 95% confidence level.

The index ? Second class correspondence is an indicator of the degree to which the duration (speed) of a first sequence of states coincides with the duration (speed) of a second sequence of states having a similar length (coincidence in speed or time). The index ? it is determined by the following expression: [[[[ENTER PAGE. 22]]]] where (t _. +? - t-)? is the time interval between state i and state i + 1 in the first sequence, (ti +? -ti) 2 is the time interval between state i and state i + 1 in the second sequence; "< -" is the minus module operator, where the module is 24 for time intervals measured in hours and 60 for time intervals measured in minutes; and "T" is the operator modulo plus, where the module is 24 par? i time intervals measured in hours and 60 for time intervals measured in minutes. The third-class correspondence index F is an indicator of the degree to which the frequency of a first sequence of states coincides with the frequency of a second sequence of states having a similar length (frequency or period correspondence). Referring to figure 12, which illustrates how the user moves through the different states (indicated on the vertical axis) as a function of time (indicated on the horizontal axis), the index F of third correspondence Class is determined as follows. The six movement cycles represented in figure 12 can be interpreted as recurrent with different frequency Fl, F2, where Fl is the frequency of the two largest movement cycles and F2 is the frequency of the four shorter movement cycles. The frequency Fk of an MC or MT is determined by the following expression: [[[[ENTER THE FIRST EQUATION OF PAGE. 23] j]] where n = p + 1. The frequency Fk 'of a sequence of states that enters is determined by the following expression: [[[[ENTER THE SECOND EQUATION OF PAGE. 23]]]] as seen in Figure 12. In this way, the index F of third-class correspondence is determined by the following expression: [[[[ENTER THIRD EQUATION OF PAGE. 23]]]] where F = 0 indicates the exact match. It should also be recognized that the index F of third-class correspondence is determined by the following expression: [[[[ENTR ^ THE FOURTH EQUATION OF PAGE. 23]]]] where Fi is the frequency of one of the first sequences that coincided and F2 is the frequency of the other sequence that coincided. Now, in relation to Figure 11, if only one of the stored MCs or MTs has met the correspondence requirements μ with the entering sequence, (μ = al) this is provided as the prediction made by the MMP. If more than one MC or MT has met the correspondence requirements μ, then second class matching rates will be examined. If only an MC or MT stored has complied with the correspondence requirement? with the sequence that enters (? >; a3), the MMP provides it as the prediction. If more than one MC or MT has complied with the correspondence requirement, then the third class correspondence rates are examined. If only one stored MC or MT has an F that matches the F of the entering sequence, the MMP provides it as the prediction. If more than one MC or MT has an F that corresponds, then the sequence that has a constitutional constraint state with the highest priority parameter p is the one that provides the MMP as the prediction. The constitutional restrictive states that are used in correspondence processes are based on the physical construction of communication systems, which is known as the a priori system. If the MMP is instrumented in a mobile station, this information of the physical construction can be provided to the mobile through the raised messages that are sent in a control channel. The bases for restrictive states are illustrated in Figures 13a, 13b. As indicated in Figure 13a, a user of a mobile located in a given cell, say cell 0, can move to any of the six adjacent cells 1-6 in the same plane and to adjacent cells that can be up and down of this plane; in this way, in the absence of any other information, each of the eight adjacent cells can be the next state with a probability of 1/8 if the user moves randomly and uniformly in all directions. Figure 13b exemplifies a physical configuration in which a door is at one end of a corridor that is also attached to another corridor. The communication system would know a priori that the mobile user who is in one of the corridors can not pass through a wall of the corridors and enter other cells. This information can be used to identify restrictive states. The process carried out by the applicants' MMP mobile motion predictor can now be summarized by means of the following pseudocode, which is illustrated in the flow chart of Figure 14. BEGIN 1) IF an S ", f entering in a new state, DO steps 2), 3), 4), ELSE DO steps 5), 6), 7), 8), ENDIF; 2) FOR each new Sk, t that enters keep a wait queue of k states (where 1 = k = N) in an order PEPS and mark Sk, t as "state limit" based on the criterion; 3) IF the length of PDL.U. is greater than unity, and Sk, t matches μ with the first state of PDout / PDut = Püut 'Sk, t GOTO step 9); ENDIF; 4) execute the motion prediction method (MMP); 5) IF Sk, t - = Sk, .- t (for t = ts mark Sk (t as "steady state" based on the criterion; ENDIF; 6) execute the motion track detection (MTD) method; 7) execute the motion cycle detection (MCD) method; 8) keep a number j of (MC MT) in the order of replacement LRU in an IPB, (where 0 = j = M); 9) repeat from step 1); END. Using the invention of the applicants, the mobile terminals will be more intelligent to manage the use of the reservation memory and the pre-search for the mobile data and file communication systems; the administration of information, for example, by selecting the forms of transmission of information, etc .; and the use and administration of the network, for example, by balancing network traffic and assigning dynamic channels, etc. Mobile data communication will be more transparent to the user and improve the quality of the service. The functioning of the MMP of the applicants was simulated and the result of the simulation will now be described with reference to figure 15, which shows an example of the standardized results of a simulation of the MMP of the applicants in which the number of states was 100, the length of the waiting queue of the states was 500, the size of the IPB was 500, and? 0 was 0.05. The operation of the MMP was simulated over a period of 5 weeks. Figure 15 shows the relationship between the prediction ratio PR and a random factor. The random factor refers to the proportion of a movement that is only due to chance; a unit random factor means that a particular movement, or transition between states, was completely random. The prediction ratio is the relationship between the number of correctly predicted states and the total number of input states; a unit prediction ratio means that each prediction of the MMP status is correct. In Figure 15, the "optimal" line is the best result expected (theoretical), ie, if there is a regularity factor X in a movement (ie, the random factor is 1 - X), then the ratio of prediction is X. It can be seen, from figure 15, that the results of the simulated MMP followed very well the track of the optimal line; The dotted line in Figure 15 shows the results of the quadratic mean in the simulation. The efficiency of the prediction of the simulated MMP, which is the ratio of the prediction ratio between the regularity factor was about 95-¿. In carrying out the simulation some simplification considerations were made that do not necessarily manifest the conditions of a situation in the real world. In particular, the probability of the movement of a mobile from one AI to another AI was considered as a uniform distribution; in other words, constitutional restrictive states were not used (the restrictive exits were 0). Also, the time intervals between the consecutive states were considered as a Poisson distribution, which was adjusted by a daily mobility factor according to the following relationship:? =? 0 • MF where? 0 was the unadjusted density of the Poisson distribution that was assumed. The daily mobility factor MF took one of three values, according to the simulated day time: MM = 2 for times between 2000 hours and 600 hours; MF = 4 for times between 830 hours and 1600 hours; and MF = 8 for times between 600 hours and 830 hours and between 1600 hours and 2000 hours. It is considered that this behavior of the mobility factor is very close to the behaviors in the real world. The applicant's mobile motion predictor can be used in an aggressive mobility management scheme that can be called predictive mobility management. The MMP predicts where a user will be based on the user's historical movement patterns, and the data and / or services are pre-connected and / or pre-assigned to the predicted location before the user requests them. In this way, the user can access their data and / or services in the predicted place with almost the same efficiency as in the previous location. In order to distribute the network services and resources more closely to the mobile user, that is, to provide mobility of the service and resources in wireless data networks, a mobile floating agent (Mobile Floating Agent, MFA) and a mobile agent can be instrumented. mobile platform agent of the distributed system (Mobile Distributed System Platform Agent, MDSP) with the MMP of the applicants. The mobility of the service is defined as the mobility of the different logics of the service in the underlying network to satisfy the quality of the service requirements of the mobile user. Mobility of resources is defined as the mobility of resources, such as data / system programs, user data, user programs, etc., in the underlying network to meet the quality of service requirements to the mobile user. In order to manage the mobility of the service and the resources, the management of the mobility developed from the mobility of the user and the terminal is required. To support mobility efficiently, each user and each terminal can be advantageously represented in the network through the respective agents, which contain all the logic and data of the service related to the user or the terminal and control all sessions of the user or the terminal. Users and terminals are connected to access the nodes of the network, and agents provide their services in the servodrives. In networks, such as in the GSM network in Europe, the controllers of the base station act as the access nodes, and in the MSC, with its integrated registrar for the visitor's location, it acts as a server and as a location that you visit The different aspects of these intelligent networks and agents are described in L. Soderberg, "Evolvinc an Inteilígent Architectute for Personal Telecomm nication", Ericsson Review, vol. 70, no. 4, pp. 156-171 (1993), which is incorporated herein by reference. In relation to Figure 16, the MFA can be implemented as a process or series of processes, running on hosts, or routers, determined, remote, which communicate and pre-connect with local host resources and which handle a data backup memory. repeated class, variable in the name of an MDSPA. An MDSPA can be implemented as a process or series of processes, running on a determined host or router, domestic, that communicates and assigns an MFA with determined, remote hosts or routers on behalf of its mobile client user. A Distributed System Mobile Platform (MDSP) and the MFA are designed to cope with the diversity of bandwidths and connectivity of different communication links in different locations and to support mobility in the service and the resources. The MDSP usually includes functions of Information Management susceptible to localization (LSIM) and functions of Predictive Mobility Management (PMM) to support different applications, such as mobile file systems, mobile smart networks, etc. In summary, LSIM functions include information related to services or resources (in which hardware and software resources are available, network connectivity, types of communication protocols available, etc.) provided by the systems or networks in a defined geographic area. The functions of the PMM involve the predictions of the location of the mobile terminal and the functions of Assignment of the floating agent, distributed-viduals (FAA), which assign the agent to different locations according to the location predictions and provide the pre-connection to the service and mobility of the service / resource.

The LaIM in the MDSP manages the information susceptible to localization and uses maps for the different services it offers through the mobile infrastructure in different geographical locations. In addition, the LSIM informs the applications and agents in the mobile network that supports applications about changes in the location of a mobile terminal and provides the dynamic connections of the service. For example, suppose that a network has a distributed file system and several servers placed in different geographic areas. If a mobile terminal moved from a location close to one of the servers to a location close to another server, the LSIM would inform the second server and the backup memory manager of the mobile terminal that the second server is the closest: > , it may be necessary to search for a file. With the support of the MDSPA and MFA, the service logic and local resources are separated from the L network and can be moved by following their mobile users. Moreover, by using the predictive mobility management functions available with the MMP of the applicants, the logic and resources of the service can be pre-assigned and pre-connected to the places the user is moving.

Likewise, the MMP of the applicants can be used for hands-free connection and the reselection of more efficient cells. In general, a mobile phone, although accommodated in a certain cell, maintains a priority list of information concerning the adjacent cells in which it can be housed. The mobile obtains the information from the priority list by scanning the possible control channels for these adjacent cells. The MMP of the applicants can reduce the number of control channels to be scanned, and reduce the amount of information in the priority list so that only those cells that were very likely candidates for the user to move in are scanned. Of course, it is possible to modify the invention in specific forms different from those described in the foregoing without departing from the spirit of the invention. The modalities described in the above are merely illustrative and in no sense should be considered as restrictive. The scope of the invention is determined by the following claims, rather than the foregoing description, and all variations and equivalents that fall within the scope of the claims should be contained herein.

Claims (38)

1. A method for predicting the next location of a mobile terminal based on the previous stored locations of the mobile terminal, which consists of the steps: the comparison of a current sequence that includes the current location of the mobile terminal and a plurality of previous locations of the mobile terminal with each of a plurality of stored sequences, each of which includes the previous locations of the mobile terminal; the selection of one of the stored sequences based on at least one quantitative measure of a degree of coincidence between the current sequence and each of the stored sequences; and predicting the next location of the mobile terminal based on the sequence that was selected from the stored sequences.

The method of claim 1, wherein the plurality of stored sequences includes movement cycles and movement tracks.

The method of claim 1, wherein one of the stored sequences is selected based on the ratio of a number of locations in the current sequence, which are the same as the locations in a stored sequence, and the total amount of locations in the current sequence.

The method of claim 3, wherein one of the stored sequences is further selected based on the quantitative measure of the degree to which a duration of the current sequence matches a duration of each of the stored sequences.

The method of claim 4, wherein one of the stored sequences is further selected based on a quantitative measure of the degree to which a frequency of the current sequence matches a frequency of each of the stored sequences.

6. An apparatus for predicting a next location of a mobile terminal, based on the previous locations of the mobile terminal, consisting of: a memory for storing the sequences of the previous locations of the mobile terminal; the means, in communication with the memory, for comparing a current sequence, which includes the current location of the mobile terminal and a plurality of previous locations of the mobile terminal with each of a plurality of stored sequences; means for selecting one of the stored sequences based on at least one quantitative measure of the degree of coincidence between the current sequence and each of the stored sequences; and the means for generating a prediction of the next location of the mobile terminal based on the selected sequence of the stored sequences.

The apparatus of claim 7, wherein the plurality of stored sequences includes movement cycles and movement tracks.

The apparatus of claim 6, wherein the selection means includes the means for determining a ratio between the number of locations in the current sequence, which are the same as the locations in a stored sequence, and the total number of locations in the current sequence, and the sequence that was selected from the sequences stored based on the relationship.

The apparatus of claim 8, wherein the selection means further includes means for generating a second quantitative measure of the degree to which a duration of the current sequence matches the duration of each of the stored sequences, and the The sequence that is selected from the stored sequences is selected based on the ratio and the second quantitative measure.

The apparatus of claim 9, wherein the selection means further includes the means for generating a third quantitative measure of the degree to which a frequency of the current sequence coincides with a frequency of each of the stored sequences, and one of The stored sequences are selected based on the relationship, the second quantitative measure and the third quantitative measure.

11. A method for predicting the movements of a mobile terminal, consisting of the steps: (a) comparing a current sequence that includes the current location of the mobile terminal and a plurality of previous locations of the mobile terminal with each of a plurality of stored sequences, each of which includes the previous locations of the terminal; (b) determining at least one quantitative measure of the degree of coincidence between the current sequence and each of the stored sequences; and (c) if at least one of the quantitative measures exceeds a predetermined value, the use of the locations of the respective stored sequence as predictions of the movements of the mobile terminal.

The method of claim 11, wherein the quantitative measure is the ratio of the number of locations in the current sequence, which are the same as the locations in a stored sequence, and the total number of locations in the current sequence .

The method of claim 12, wherein step (b) further determines a first degree in which a duration of the current sequence matches a duration of the stored sequences, and step (c) uses as predictions of the movements of the mobile terminal the locations of the stored sequence for which the ratio exceeds a first predetermined value and the first degree exceeds a second predetermined value.

The method of claim 13, wherein step (b) further determines a second degree in which a frequency of the current sequence matches a frequency of each of the stored sequences, and step (c) uses, as predictions of the movements of the mobile terminal, the locations of the stored sequence for which the second degree exceeds a third predetermined value.

15. A method for determining normal patterns in the movements of a mobile terminal, the method consists of the steps: (a) comparing a current location of the mobile terminal with each of a plurality of previous locations stored in a queue waiting for a plurality of the previous locations, the previous locations are stored in the waiting queue in the order first inputs-first outputs; (b) if the current location matches one of the plurality of the previous locations stored in the queue, the marking of a sequence of locations consisting of: the current location, the previous location that matches the current location, and the previous locations that occurred after the previous location that matches the current location; (c) comparing the labeled sequence with each of a plurality of stored sequences and determining at least one quantitative measure of a degree of coincidence between the labeled sequence and each of the stored sequences; and (d) if the at least one quantitative measure exceeds a predetermined value, the increase, by a predetermined amount, of a priority parameter of the respective current sequence.

16. The method of claim 15 further comprises the steps: (e) e .. storing the current location of the mobile terminal in the queue, in the order of appearance of first inputs-first outputs; (f) the determination of, if the current location is at least a steady state or a limit state; and performing steps (a) - (d) if the current location is at least a stationary state or a limit state.

The method of claim 16, wherein step (c) determines a ratio of the number of locations in the marked sequence, which are the same as the locations in a stored sequence, and the total number of locations in the sequence. marked sequence.

The method of claim 17, wherein step (c) further determines a degree in which a duration of the tagged sequence coincides with a duration of each of the stored sequences.

The method of claim 18, wherein step (c) further determines a degree in which a frequency of the tagged sequence matches a frequency of each of the stored sequences.

20. An apparatus for determining the normal patterns of the movements of a mobile terminal, the apparatus consists of: a memory for storing a waiting queue of a plurality. from previous locations of the mobile terminal, the previous locations are stored in the waiting queue in the order first inputs-first outputs; the first means for comparing the current location of the mobile terminal with each of the plurality of the previous locations stored in the waiting queue; the means to mark a sequence of locations, the medium consists of: the current location, the previous locations that coincide with the current location, and the previous locations that were presented after the previous location that coincides with the location. If the current location matches one of the plurality of previous locations stored in the wait queue. A second means for comparing the tagged sequence with each of a plurality of sequences of stored locations and for determining at least one quantitative measure of a degree of match between the tagged sequence and each of the stored sequences; the method for increasing a priority parameter of the respective stored sequence by a predetermined amount when the "at least one quantitative measure" exceeds a predetermined value.

21. The apparatus of claim 20, wherein the current location of the mobile terminal is stored in the waiting queue in the order of appearance of first inputs-first outputs, and furthermore consists of the means to determine whether the current location is at least a steady state or a limit state.

22. The apparatus of claim 21, wherein the second means determines a relationship between the number of locations in the tagged sequence, which are the same as the locations in a stored sequence, and the total number of locations in the tagged sequence.

23. The apparatus of claim 22, wherein the second means further determines a degree in which a duration of the marked sequence coincides with a duration of each of the stored sequences.

The apparatus of claim 23, wherein the second means further determines a degree in which a frequency coincides with a frequency of each of the stored sequences.

25. A method for determining the normal patterns of movements of a mobile terminal, the method consists of the steps: (a) determining whether a current location of the mobile terminal is at least a steady state or a limit state; (b) mark a sequence of locations containing the current location, one of the most recent previous stationary states and one of the most recent previous limit states, and the previous locations that occurred between the most recent previous steady state and the limit state previous most recent (c) comparing the labeled sequence with each of a plurality of stored location sequences and determining at least one quantitative measure of a degree of match between the labeled sequence and each of the stored sequences; and d) if the "at least one quantitative measure" exceeds a predetermined value, the increment of a priority parameter of the respective stored sequence by a predetermined amount.

26. The method of claim 25 further comprises the steps of: storing the current location in a queue from a plurality of previous locations, the locations are stored in the queue in the order first first innings-out, wherein steps (a) - (d) are carried out if the current location is at least a stationary state or a limit state.

27. The method of claim 25, wherein the ratio of a number of locations in the tagged sequence is determined, which is the same as the locations in a stored sequence, and the total number of locations in the tagged sequence.

The method of claim 27, wherein a degree is determined in which a duration of the marked sequence coincides with a duration of each of the stored sequences.

29. The method of claim 28, wherein a degree is determined in which a frequency of the marked sequence coincides with a frequency of each of the sequence? stored.

30. An apparatus for determining normal patterns in the grinding of a mobile terminal, the apparatus consists of: means for determining whether a current location of the mobile terminal 1 is at least in a steady state or a limit state; the means to mark a sequence of locations containing: the current location, the most recent previous steady state or the most recent previous limit state, and the previous locations that occurred between the most recent previous steady state or the most recent previous limit state; the meldium for comparing the labeled sequence with each of a plurality of sequences of stored locations and for determining at least one quantitative measure of a degree of coincidence between the labeled sequence and each of the stored sequences; and means for increasing a priority parameter of the respective stored sequence by a predetermined amount if the "at least one quantitative measure" exceeds a predetermined value.

31. The apparatus of claim 30 further comprises a minority for storing the current location in a waiting queue of a plurality of previous locations, the locations being stored in the queue in the order first inputs-first outputs.

32. The apparatus of claim 31, wherein the means for comparison and determination determines a ratio of the number of locations in the marked sequence, which are the same as the locations in a stored sequence, and the total number of locations in the sequence. marked sequence.

The apparatus of claim 32, wherein the means for comparing and determining determines a degree in which a duration of the marked sequence coincides with a duration of each of the stored sequences.

34. The apparatus of claim 33, wherein the means for comparing and determining determines a degree in which a frequency of the marked sequence coincides with a frequency of each of the stored sequences.

35. A communication network consisting of: a plurality of servers, the servers are placed in respective geographic areas and organized in a distributed file system; a mobile terminal having the means to communicate with the server closest to the mobile terminal, wherein the communication means has access to the application files and to the data files stored in the servers; and a mobile platform of the distributed system that has the means to control the distributed file system of the servers and the means to predict the next location of a mobile terminal, where the control means distributes between the servers the information susceptible to location with base in a next location predicted by the prediction medium.

36. The communication network of claim 35, wherein the MDSP includes at least one MDSP agent and at least one mobile flotation agent (MFA), the MFA is associated with the mobile terminal, and the MDSP agent is communicates and preassigns the MFA to at least one of the servers on behalf of the mobile terminal based on the next location predicted by the prediction means.

37. The communication network of claim 36, wherein the MFA is a process that is executed on a remote server from the mobile terminal that communicates and pre-connects with a local server to the mobile station and handles a repeated data cache. on behalf of the MDSP agent.

38. A cellular radiotelephone system having a plurality of base stations and a mobile terminal, each of the base stations transmits the respective control information in a respective control channel, an apparatus for prioritizing the base stations for a connection between the mobile terminal and a base station, the system consists of: the means for predicting the next locations of the mobile terminal based on the previous locations of the mobile terminal L, where the means to predict includes: a memory for storing the sequences of the previous locations of the mobile terminal; the means, in communication with the memory, for comparing a current sequence, the means includes the current location of the mobile terminal and a plurality of previous locations of the mobile terminal with each of a plurality of stored sequences; the means for selecting one of the stored sequences based on at least one quantitative measure of a degree of coincidence between the current sequence and each of the stored sequences; and the means for generating predictions of the next locations of the mobile terminal based on the selected sequence of the stored sequences; and means for sweeping the control channels of a plurality of base stations and for maintaining a priority list of information related to the swept control channels; where the means for the sweep makes the exploration of the control channels based on the predictions generated by the prediction means. SUMMARY OF THE INVENTION The present invention relates to methods and apparatus for detecting and predicting movement patterns of mobile radio transceivers, such as mobile cellular telephones. A method for predicting a next location of a mobile terminal with base includes the step of comparing a current sequence, which includes the current location of the mobile terminal and a plurality of previous locations of the mobile terminal with each of a plurality of sequences stored, which each includes the previous locations of the mobile terminal. The method also includes the steps of selecting one of the sequences stored based on at least one quantitative measure of the degree of coincidence between the current sequence and each of the stored sequences, and the prediction of the next location of the mobile terminal based on the selected sequence of the stored sequences. Also described are methods and apparatus for determining normal patterns in the movements of a mobile terminal, as well as a communication network with several servers, wherein a mobile terminal has a device for communicating with the nearest server. The device has access to the application and data files stored on the servers. It also describes a mobile platform of the distributed system that has a device for controlling the distributed file system, but also a device for predicting the next location of a mobile terminal, where this device distributes, among the servers, information related to the location on, based on a predicted next location.