CN1747359A - Self-adaptive rate control based on load in mobile telecommunication system - Google Patents

Abstract

The invention can efficiently assign and dispatch resources in real time between many service requests to improve speed control method and utilization of system resources.

Description

In the mobile communication system based on the adaptation rate control method of load

Technical field

The present invention relates to the method for rate control in a kind of mobile communication system, especially in the mobile communication system of supporting the several data transmission rate, be between the service request under the different transmission conditions and distribute transmission rate, thereby when satisfying each requested service quality, improve the whole service efficiency of system.Speed realization reasonable resources distribution and scheduling that rate controlled is distributed by being adjusted into different user, therefore, rate controlled also is a kind of implementation method of scheduling strategy.

Background technology

Flourish along with mobile communication, various new business emerge in an endless stream, and wherein the development of data service is particularly rapid.And the limitation of radio spectrum resources and communication environments has influenced the raising of data traffic transmission speed to a certain extent.Therefore, the research focus that how to make full use of system resource, provides high as far as possible transmission rate to become moving communicating field for the user.The proposition of rate control techniques has well solved this problem.This method according to the current transmission rate of each service request, system resource current utilize situation, can the transmission rate of each service be carried out in time, effectively adjust, thereby system resource is fully used in conjunction with specific probability system of selection.

Two important bases for estimation are in the method for rate control: system resource utilization situation and probability system of selection.The criterion of utilization of resources situation and the value of probability all can produce material impact to the result of rate controlled.The method of rate control criterion that must use simply, easily measure and reasonably probability value efficiently.Common rate control techniques adopts parameters such as system's available horsepower, system load and system's residual capacity to weigh as the standard of system resource utilization situation, carries out the speed adjustment based on fixation probability.The method of adjustment of fixation probability is easy to realize, but the variation of adaptive system utilization of resources situation dynamically.

Summary of the invention

For the tracking system loading condition, and adjust the rate controlled probability in real time to adapt to the variation of utilization of resources situation, the inventor has proposed a kind of method of rate control based on system load, adaptive adjustment speed lifting probability.Its basic principle is that judge according to the load value of current system should raising speed rate or the transmission of reduction of speed rate, and calculate the probability of raising speed rate or reduction of speed rate according to load-threshold value, probability is adjusted speed according to this, and the maximum rate that provides of frame of reference available horsepower, determines the transmission rate of following one-period.

The concrete performing step of the present invention is:

Step 1: the load value of adding up each controlling party at network side.

Step 2: determine the overload indication bit of each controlling party according to the load value in the step 1, whether the receiver that indication bit has indicated current base station transships, and as overload, then requires the power from user side that it received to reduce, otherwise can improve.

Step 3: determine the speed adjustment strategy that user side should be taked according to the overload indication bit of each controlling party in the controlling party set (abbreviation control sets) of user side association in this controlling party: raising speed rate or reduction of speed rate.

Step 4: adjust strategy and load-threshold value according to the speed of determining in the load value of adding up in the step 1, the step 3, calculate the tactful probability value of user side speed accent that each controlling party is controlled, i.e. the probability lower limit of the probability upper limit of raising speed rate or reduction of speed rate.

Step 5: overload indication bit and speed according to each controlling party in the user side control sets are adjusted probability, determine the rate value of method of rate control decision.

Step 6: calculate the maximum rate value that each user side allows transmission according to system's available horsepower.

Step 7: select rate value less in step 5 and the step 6 as the final transmission rate of determining.

After the next rate controlled cycle began, repeating step 1 was to step 7.

Determine that according to the different loading condition of each controlling party speed adjusts strategy in the top step 3, and in step 4, determine in real time that according to current load value and load-threshold value speed adjusts the limiting value of probability.What calculate for the raising speed rate is the probability upper limit, if the probability of raising speed rate, then can cause this controlling party overload greater than this probability upper limit; What calculate for the reduction of speed rate is the probability lower limit, if the probability of reduction of speed rate, equally also can cause this controlling party overload less than this probability lower limit.Step 7 takes into account the supporting factor of available horsepower to speed.The present invention measures and probability calculation by real-time, the tracking system loading condition, and in time take measures to prevent system overload and guarantee making full use of system resource.

Description of drawings

In conjunction with the following drawings and instantiation will be convenient to understand principle of the present invention, step, characteristics and advantage to the detailed description done of invention, in the accompanying drawing:

Fig. 1 is the job step of expression the present invention in the CDMA2000 example;

Fig. 2 represents the present invention in the CDMA2000 example, the curve chart and the growth ratio curve of sector average load value when sector average load value and employing fixation probability.Each sector has two voice users and several data users.Axis of abscissas is represented each sector data user number, and axis of ordinates is represented the average load and the growth of load percentage of sector.

Fig. 3 represents the present invention in the pairing CDMA2000 example of Fig. 2, the curve chart and the throughput increase percentage of sector average throughput when sector average throughput and employing fixation probability.

Embodiment

Because the implementation of data service is a lot of in the mobile communication,, using a kind of representative and blanket example at this for the ease of understanding---method of rate control proposed by the invention is set forth in CDMA2000 suggestion (being formulated by 3GPP2).The present invention generally is suitable for for the application scenario that all have essential characteristic in this example.

Use formula (1) to calculate the load contribution of user i to Active Set sector j:

${\mathrm{\η}}_{i,j}=\frac{1}{1+\frac{W}{{(E_b/N_0)}_i\×R_i\×{\mathrm{\υ}}_i}}---\left(1\right)$

Wherein, W is a system bandwidth, E _b/ N ₀Be user's bit energy and the ratio of background heat noise power spectral density, R _iBe transmission rate, υ _iBe the voice activation factor.For the data user, the voice activation factor is 1.According to " 1xEV-DVEvaluation Methodology (V13) " literary composition of 3GPP2, the voice activation factor of getting voice user is 0.406.

It is the user load contribution sum of Active Set with this sector that the load value of sector equals all:

${\mathrm{Load}}_j=\underset{j\∈\mathrm{ActiveSet}\left(i\right)}{\mathrm{\Σ}}{\mathrm{\η}}_{i,j}---\left(2\right)$

The determination methods of sector overload indication bit is:

${\mathrm{OLB}}_j=\left\{\begin{array}{cc}0& {\mathrm{Load}}_j<{\mathrm{<figure-callout\; id="1"\; label="Load\; TH"\; filenames="A20051010260100101.png,A20051010260100111.png"\; state="\{\{state\}\}">Load</figure-callout>}}_{\mathrm{TH}}\\ 1& {\mathrm{Load}}_j\&GreaterEqual;{\mathrm{Load}}_{\mathrm{TH}}\end{array}---\left(3\right)\right.$

Stipulated maximum load value, the threshold value Load of load just in " IS-2000 Release C Simulation Configuration Specification " literary composition of 3GPP2 _THBe 0.45.When sector load greater than 0.45 the time, show that current sector load is bigger, the user answers the transmission of reduction of speed rate; Otherwise, illustrating that there is surplus resources current sector, the user can the transmission of raising speed rate.But raising speed rate and reduction of speed rate will be according to certain probability, can not all users raising speed rate or reduction of speed rates simultaneously.If all users are reduction of speed rates simultaneously, can cause system resource idle suddenly, thisly be idle in next time that rate controlled can cause all user's raising speed rates again, thereby cause system overload in the cycle.And so forth, system fluctuates under idle and overload, causes the instability of system.Main method of the present invention is exactly by adjusting the probability of ascending, descending speed, making the approaching as much as possible threshold value that still is no more than of load value of system.Details are as follows for the computational methods of speed adjustment probability.

For the situation of sector j permission user raising speed rate, promptly current Load _j＜Load _TH, establishing user i present rate is R _i, the speed probability that rises to the present rate twice is p _j ^↑(0≤p _j ^↑≤ 1), keeps present rate R _iProbability be l-p _j ^↑Use speed estimation function Ev (R _i, p _j ^↑) calculate the speed of following one-period and bring the load calculation formula into.The mathematic expectaion function that uses transmission rate in the next rate controlled cycle in this example is as the speed estimation function: $\mathrm{Ev}(R_i,p_j^{\&UpArrow;})=\stackrel{\&OverBar;}{R_{\&UpArrow;}}=2\&times;R_i\&times;p_j^{\&UpArrow;}+R_i\&times;(1-p_j^{\&UpArrow;})=R_i\&times;(1+p_j^{\&UpArrow;}).$ Will

Bring the load calculation formula into, obtain

$\stackrel{\&OverBar;}{{\mathrm{Load}}_j}=\underset{j\&Element;\mathrm{ActiveSet}\left(i\right)}{\mathrm{\&Sigma;}}\frac{1}{1+\frac{W}{{(E_b/N_o)}_i\&times;(1+p_j^{\&UpArrow;})\&times;R_i\&times;{\mathrm{\&upsi;}}_i}}---\left(4\right)$

Formula (4) is variable p _j ^↑(0≤p _j ^↑≤ 1) increasing function.Find the solution satisfied $\stackrel{\&OverBar;}{{\mathrm{Load}}_j}<{\mathrm{Load}}_{\mathrm{TH}}$ Maximum probability p under the condition _j ^↑, be the maximum raising speed rate probability that makes sector j nonoverload.

For the situation of sector j requirement user reduction of speed rate, the promptly current Liad that goes up _j〉=Load _TH, establish user i speed reduced to R _i/ 2 probability is p _j ^↓(0≤p _j ^↓≤ 1), keeps present rate R _iProbability be 1-p _j ^↓The same mathematic expectaion function that uses transmission rate in the next rate controlled cycle is as the speed estimation function: $\mathrm{Ev}(R_i,p_j^{\&DownArrow;})=\stackrel{\&OverBar;}{R_{\&DownArrow;}}=R_i\&times;p_j^{\&DownArrow;}/2+R_i\&times;(1-p_j^{\&DownArrow;})=R_i\&times;(1-p_j^{\&DownArrow;}/2).$ With each user

Bring in the load calculation formula, obtain:

$\stackrel{\&OverBar;}{{\mathrm{Load}}_j}=\underset{j\&Element;\mathrm{ActiveSet}\left(i\right)}{\mathrm{\&Sigma;}}\frac{1}{1+\frac{W}{{(E_b/N_o)}_i\&times;(1-p_j^{\&DownArrow;}/2)\&times;R_i\&times;{\mathrm{\&upsi;}}_i}}---\left(5\right)$

Formula (5) is variable p _j ^↓(0≤p _j ^↓≤ 1) subtraction function.Find the solution satisfied $\stackrel{\&OverBar;}{{\mathrm{Load}}_j}<{\mathrm{Load}}_{\mathrm{TH}}$ Minimum probability value p _j ^↓As making sector j change the reduction of speed rate probability of lucky nonoverload into, thereby realize making full use of to system resource by overload.

According to " 1xEV-DV Evaluation Methodology (V13) ", can comprise three sectors in user's the Active Set at most.In order to satisfy the requirement that each activates the sector nonoverload, have only when all activated sector requires the raising speed rate, Cai the transmission of user's raising speed rate, and final raising speed rate probability is the minimum value of all raising speed rate probability.As long as there is an Active Set sector to require user's reduction of speed rate, then reduction of speed rate transmission of user, the probability of reduction of speed rate is got the maximum of all reduction of speed rate probability.Therefore, the finally definite overload indication bit of user can be expressed as its Active Set sector overload indication bit sum:

${\mathrm{OLB}}_i=\underset{j\&Element;\mathrm{ActiveSet}\left(i\right)}{\mathrm{\&Sigma;}}{\mathrm{OLB}}_j---\left(6\right)$

The probability system of selection of user i is expressed as:

$p_i^{\&UpArrow;}=\underset{\mathrm{RA}{B}_j=0}{\underset{j\&Element;\mathrm{ActiveSet}\left(i\right)}{\min }{p}_j^{\&UpArrow;}},({\mathrm{OLB}}_i=0)---\left(7\right)$

$p_i^{\&DownArrow;}=\underset{\mathrm{RA}{B}_j=1}{\underset{j\&Element;\mathrm{ActiveSet}\left(i\right)}{\max }}{p}_j^{\&DownArrow;},({\mathrm{OLB}}_i=1)$

Produce the random number in [0, the 1] scope that is evenly distributed at random, if this random number is less than p _i ^↑(p _i ^↑), then the user is 2 times (1/2 times) of present rate with the speed rising (decline) of next rate controlled in the cycle.

In addition, also considered of the restriction of system's available horsepower among the present invention to speed.Select speed less in the transmission rate that supported peak transfer rate of current available horsepower and rate controlled obtain as final speed.

As can be seen, what the present invention proposed adjusts the adaptive algorithm of speed lifting probability in real time at system loading conditions, has following characteristics more than comprehensive:

1. be purpose with the whole utilization ratio that improves system resource, for data service, promptly obtain bigger throughput of system.

2. avoid the generation of system overload phenomenon, also guarantee the stability of system simultaneously.

3. the consideration high real-time along with the continuous variation of system status, is adjusted speed lifting probability in real time, avoids the blindness under the fixation probability situation.

4. insensitive to concrete implementation and protocol specification, possess compatibility.

In a word, the present invention is a kind of resource allocation and dispatching technique of flexible practicality, its objective is the efficient of adjusting with resource allocation and speed in the method raising mobile communication of simple structure and easily realization.

Claims (13)

1. one kind is used for the method for rate control that mobile communication resources distributes and dispatches, based on the communication equipment that comprises in the system, the communication resource of available launch power and communications band etc., at the communication channel of dividing and making up and constitute by the difference of each communication resource, network controls side is a decision-making foundation with the state information of system's essential information and user side, these state informations comprise: system bandwidth, user side transmission rate and channel conditions, activity factor, and then, adjust and distribute transmission rate dynamically between each user according to the optimized criterion of indexs such as service quality that system resource utilization efficient and user side are obtained of system design goal defined.In this process, adopted the adaptive probability system of selection, by probability according to present load situation and load-threshold value calculating, make the probability of the real-time adjustment rate variation of user side, thereby the variation of accurate tracking system load state, and make corresponding adjustment to adapt to load variations, make system load be tending towards threshold value but and be no more than threshold value, that is to say under the prerequisite of system's nonoverload, make full use of system resource.The basic performing step of this method is:

The load of computing network side controlling party, and this load compared with load-threshold, set the overload indication bit of this controlling party according to comparative result: load is greater than load-threshold, the overload indication bit is set is in overload to show current system, load is provided with the overload indication bit and shows that still there is surplus resources in current system less than load-threshold;

Determine that according to the overload indication bit user side speed that controlling party controls adjusts trend: if the overload indication bit indicates system then raising speed rate of idling-resource is arranged still, on the contrary reduction of speed rate then;

Calculate the speed that user side should adopt in this sector according to present load, load-threshold, speed adjustment trend and speed estimation function and adjust probability;

For each user,, determine the overload indication bit and the speed adjustment probability of user side according to the integrated status of its one or more network controls sides.In order to guarantee all nonoverloads of each controlling party, the probability of raising speed rate is the minimum value in the controlling party raising speed rate probability, and the probability of reduction of speed rate is the maximum in the controlling party reduction of speed rate probability.

The liter that user side is determined in rapid according to its overload indication bit and previous step or the probability of reduction of speed rate are determined its transmission rate in next control cycle.

Speed less in the transmission rate that user side peak transfer rate that the selective system available horsepower allows and method of rate control are determined is as final transmission rate.

2. described in claim 1, be used for that mobile communication resources distributes and the method for rate control of scheduling, it is characterized in that: the allocation schedule that the system's essential information of network side institute foundation and the state information of user side have comprised resource is here handled and all possible information of method of rate control institute foundation, and the generation of these information and obtaining.

3. described in claim 1, be used for that mobile communication resources distributes and the method for rate control of scheduling, wherein, the information spinner of rate controlled decision-making institute foundation will comprise: the information of the dynamic change of system mode and user side wireless propagation environment etc., the former comprises system bandwidth, load-threshold value and the load statistics information that obtained etc., and the latter comprises that each user side channel conditions is measured and the measurement of transmission rate and prediction etc.

4. described in claim 1, be used for that mobile communication resources distributes and the method for rate control of scheduling, wherein, user side is characterised in that the service request that representative is vied each other, they dispatch and are assigned with the transmission rate that corresponding system resources allows by the rate controlled side that grasps scheduling of resource, in this process, network side is according to making the indexs such as service quality that the level of resources utilization and user side obtain in the system reach optimized criterion, the dynamic probability of adjusting the user side rate variation, user side is according to this probability and residing situation thereof, comprise each controlling party state and oneself state thereof, adjust its transmission rate.

5. described in claim 1, be used for that mobile communication resources distributes and the method for rate control of scheduling, network controls side wherein, its effect be exactly to receive and the process user end report and submit to it carry out speed the time control institute's foundation for information about, and make a policy user side informed in its control information of grasping.Network side and user side all comprise method of rate control, and two parts method finally realizes rate controlled by the communication of certain way.

6. described in claim 1, be used for that mobile communication resources distributes and the method for rate control of scheduling, it is characterized in that: setting network side controlling party is calculated the dynamic criterion of adjusting probability, this criterion comprises the calculating of dynamic raising speed rate probability and reduction of speed rate probability, by set dynamic criterion avoid user side simultaneously raising speed rate or reduction of speed rate cause the rapid variation of system load, and the chain reaction that causes thus is to the adverse effect of the stability of a system.

7. described in claim 1, be used for that mobile communication resources distributes and the method for rate control of scheduling, it is characterized in that: the speed of setting user side is adjusted criterion, this criterion comprises determining of user side overload indication bit and corresponding speed adjustment probability, and guaranteed under the prerequisite that satisfies each controlling party nonoverload, to utilize system resource fully by the probability selection criterion.

8. the probability calculation that belongs in the dynamic criterion in network controls side described in claim 6 is characterised in that: use the speed estimation function of user side in the next cycle, the user side transmission rate when replacing probability calculation in the employed load calculation formula.

9. the raising speed rate probability calculation that belongs in the dynamic criterion in network controls side described in claim 6 is characterised in that: under the state of controlling party and nonoverload, improve the transmission rate of user side with the highest probability, the load of realization network side increases also approaching but is no more than load-threshold, thereby realizes making full use of Internet resources.

10. the reduction of speed rate probability calculation that belongs in the dynamic criterion in network controls side described in claim 6 is characterised in that: under the state that controlling party has been transshipped, reduce the transmission rate of user side with the probability of minimum, make network side load near and be lower than load-threshold, when solving network side overload phenomenon, realize making full use of to Internet resources.

11. the user side speed that belongs to described in claim 7 is adjusted the speed of criterion and is adjusted probability and select to be characterised in that: a plurality of network controls side calculate the adjustment probability in, the load of selecting to satisfy this equal nonoverload of user side all-network side controlling party and making each controlling party is all near the probability of load-threshold.When user side overload indication bit indication user raising speed rate, minimum probability in the raising speed rate probability that the controlling party of selection permission user raising speed rate is calculated; When user side overload indication bit indication user reduction of speed rate, select maximum probability in the reduction of speed rate probability that all controlling parties that require user's reduction of speed rate calculate.

12. the speed estimation function that belongs to described in claim 8 is characterised in that: can be might transmission rate mathematic expectaion, perhaps other can characterize the estimation function of user side speed characteristic.

13. the method for rate control that is used for mobile communication resources distribution and scheduling described in claim 1 is characterized in that its basic performing step is:

The load of computing network side controlling party, and this load compared with load-threshold, set the overload indication bit of this controlling party according to comparative result: load is greater than load-threshold, the overload indication bit is set is in overload to show current system, load is provided with the overload indication bit and shows that still there is surplus resources in current system less than load-threshold;

Determine that according to the overload indication bit user side speed that controlling party controls adjusts trend: if the overload indication bit indicates system then raising speed rate of idling-resource is arranged still, on the contrary reduction of speed rate then;

Calculate the speed that user side should adopt in this sector according to present load, load-threshold, speed adjustment trend and speed estimation function and adjust probability;

For each user,, determine the overload indication bit and the speed adjustment probability of user side according to the integrated status of its one or more network controls sides.In order to guarantee all nonoverloads of each controlling party, the probability of raising speed rate is the minimum value in all raising speed rate probability, and the probability of reduction of speed rate is the maximum in all reduction of speed rate probability.

The liter that user side is determined in rapid according to its overload indication bit and previous step or the probability of reduction of speed rate are determined its transmission rate in next control cycle.

Less speed is as final transmission rate in the transmission rate that the user side peak transfer rate that selective system can power be supported and method of rate control are determined.

Images