GB2493215A - Cognitive radio sensor nodes with plural detectors - Google Patents

Abstract

A cognitive radio system is provided comprising a master node (fig.6) and a plurality of sensor nodes 5. One or more of the sensor nodes has a plurality of detector modules 349-n that can detect the presence of a primary user transmission using a number of different sensing techniques. Configuration parameters are determined for the different detector modules in advance for the different types of primary signal that the sensor node may be required to detect. The plurality of detector modules operate in parallel and a decision is taken that there is a primary signal transmission if any of the detector modules detects the presence of the primary signal. The detectors may detect signal energy, signal power or cyclo-stationary features of the signal.

Description

Communication System The present invention relates to a communications system and to parts and methods thereof. The invention has particular, although not exclusive relevance to cognitive radio systems and devices thereof.

In wireless communications, traditionally, licensing gave systems exclusive access to distinct blocks of the electromagnetic spectrum. This is a good way to eliminate the danger of harmful interference but it can leave the majority of the allocated block of spectrum underused when and where the license holder is not active.

In order to use spectrum holes (areas in which licensed users of the spectrum are not active for a given time), the idea of opportunistic access to spectrum was suggested. This idea involves two different kinds of spectrum users: legitimate users (or primary licensed users), and secondary users (or cognitive users) which are allowed to use a part of the spectrum only if the primary users are not using it or only if the generated interference (to the primary users) is acceptable.

A Cognitive Radio Device has been defined as a terminal which is aware of its electro-magnetic environment and able to adapt its transmission accordingly. A Cognitive Radio Device may sense a specific frequency band in order to estimate its occupancy and take a decision whether it transmits or not. However, the primary signals to be detected are subject to different environment changes (shadowing, fading, path loss) and therefore sometimes primary licensed users are very difficult to detect. Existing systems have been proposed that aim to maximise the probability of the device detecting the primary user transmission.

However, these techniques also increase the probability of false detection -i.e falsely detecting the presence of the primary user when the primary user is in fact not present. If the sensor nodes have a relatively high probability of false detection then it becomes very difficult for the secondary users to use the band effectively -as they are always stopping their transmissions unnecessarily due to false detections of the primary user.

The present invention aims to improve the sensing accuracy of a single cognitive radio device that balances the competing requirements of maximising the probability of detecting the primary user whilst minimising the probability of false detections.

According to one aspect, the invention provides a communication node for use in a cognitive radio communications system, the communication node comprising: transceiver circuitry for transmitting signals to and for receiving signals from at least one other communication node within a desired operating frequency band; a plurality of detectors being configured to operate in parallel to process signals received by the transceiver circuitry using a respective different processing technique to determine if a primary user of the desired operating frequency band is currently transmitting within the desired operating frequency band; and a communications controller operable to control the transmissions of the communication node in dependence upon detection results made by the plurality of detectors. The communications controller may control the transmissions of the communication node directly or indirectly in response to the detection results made by the plurality of detectors. In the case of indirect control, the detection results may be sent to a remote device which then sends a control signal to the communication node for controlling its transmissions.

The communication node may also have a decision module that receives the detection results from the detectors and that decides whether or not the primary user of the desired operating band is currently transmitting within that frequency band. In this case, the decision module may decide to stop transmitting within the desired operating frequency band if any of the detectors detects the primary user transmissions. If none of the detectors detects a primary user transmission, then the decision module may decide to allow the communication node to transmit within the desired operating frequency band. The communication controller will normally control the transmissions of the communication node in dependence upon the decision of the decision module.

In one embodiment, the detectors are configurable to detect for different types of primary user transmissions and the node may additionally comprise a sensing manager that configures the detector modules in accordance with configuration data that depends upon an expected type of primary user transmission for the desired operating frequency band. The configuration data may include data relating to a respective detection threshold for each of the detectors. This data may define a target probability of false detection for the detector and each detector may determine a detection threshold that it should use in dependence upon the defined target probability of false detection. The configuration data for a detector may also include data defining a sensing duration over which the detector is to sense for the primary user transmissions.

In a preferred embodiment, the communication node includes a database of configuration data for different types of primary user transmissions and for the different detectors. In this case, the communication node may receive an indication of a primary user signal type from a remote device and will retrieve configuration data from said database corresponding to the indicated type of primary user signal.

The communication node may further comprise a noise estimator that estimates a noise level within the desired operating frequency band and wherein a subset of the detector modules use the estimated noise level when processing the signals received by the transceiver circuitry to determine if the primary user of the desired operating frequency band is currently transmitting within the desired operating frequency band.

In one embodiment, one or more of the detectors determine the energy of signals within the desired operating frequency band and one or more other detectors detect at least one cyclo-stationary feature within the received signal that indicates the presence of the primary user transmission.

The present invention also provides a communication node for use in a cognitive radio communications system, the communication node comprising: transceiver circuitry for transmitting signals to and for receiving signals from at least one other communication node; a database storing configuration data for a plurality of detectors that are used by at least one other communication node, the configuration data including, for each detector, data relating to a detection threshold to be used by the detector for each of a plurality of different types of primary user signals; and a configuration module operable to send configuration data for the detectors in dependence upon a type of primary user signal that is expected to be transmitted within a desired operating frequency band.

The data relating to a detection threshold for a detector may define a target probability of false detection and/or a sensing duration over which the detector is to sense for the primary user transmissions.

The present invention also provides a method performed by a communication node in a cognitive radio communications system, the method comprising: transmitting signals to and receiving signals from at least one other communication node within a desired operating frequency band; using a plurality of detectors that operate in parallel to process signals received by the transceiver circuitry using a respective different processing technique to determine if a primary user of the desired operating frequency band is currently transmitting within the desired operating frequency band; and controlling the transmissions of the communication node in dependence upon detection results made by said plurality of detectors.

The present invention also provides a method performed by a communication node in a cognitive radio communications system, the method comprising: transmitting signals to and receiving signals from at least one other communication node; storing a database having configuration data for a plurality of detectors that are used by at least one other communication node, the configuration data including, for each detector, data relating to a detection threshold to be used by the detector for each of a plurality of different types of primary user signals; and transmitting configuration data for the detectors in dependence upon a type of primary user signal that is expected to be transmitted within a desired operating frequency band.

The invention provides, for all methods disclosed, corresponding computer programs or computer program products for execution on corresponding equipment, the equipment itself (user equipment, nodes or components thereof) and methods of updating the equipment.

Exemplary embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings in which: Figure 1 schematically illustrates the components of a communication system to which the invention is applicable; Figure 2a is a block diagram illustrating the main components of a sensor node shown in Figure 1; Figure 2b illustrates the data maintained in the database shown in Figure 2a; Figure 3 is a schematic diagram illustrating the functional interoperation of the software modules shown in Figure 2a during a sensing period; Figure 4 is a timing diagram illustrating the parallel operation of the detectors shown in Figure 2a; Figure 5 is a flow chart illustrating the way in which a target false alarm probability is determined for different signal types and different detectors; and Figure 6 is a block diagram illustrating the main components of a master node that may be used in an alternative embodiment.

Overview Figure 1 schematically illustrates a cognitive wireless communication network 1 that has a master node 3 and one or more sensor nodes 5. The master node 3 and the sensor nodes 5 are secondary users of one or more frequency bands that are licensed to one or more primary users. The primary users will typically include wideband primary transmitter(s) 6 that transmit over the whole frequency band and narrow band primary transmitters 7 that transmit over different narrow bands within the licensed band. As an example, the licensed band may correspond to a television channel (DVB-T operating on 6MHz or 8MHz channels) and the wideband primary transmitter will transmit over this whale band whilst the narrow band primary transmitters 7 (which may be PMSE (Programme Making Special Event) devices) may transmit over narrow bands of about 200kHz within the larger 6MHz or 8MHz channels. In addition to the master node 3 and the sensor nodes 5, the secondary users of the licensed frequency band may also include other communication nodes (not shown) that do not perform any sensing but whose transmissions are controlled by the master node. The secondary users are able to communicate within all or a part of the licensed frequency band whilst the primary users 7 are not transmitting in all or that pad of the licensed frequency band; or as long as the interference caused by the transmissions of the secondary users is acceptable to the primary users 6, 7. Typically the master node 3 will be fixed in location (and may be a base station) whilst the sensor nodes 5 can move (and may be mobile telephones). In this embodiment, the sensor nodes 5 sense or detect if there are any primary user transmissions and report the results to the master node 3 which uses the results to decide if it is allowable for any of the secondary users to transmit in any part of the licensed frequency band.

As will be explained in more detail below, one or more of the sensor nodes 5 employ multiple different detectors that operate in parallel using different detection processes or algorithms to detect if there are any primary user transmissions and a decision is taken based on whether or not any of the detectors detects primary user transmissions. If none of the detectors detects the presence of a primary user transmission, then the sensor node 5 decides, for the current sensing period, that there is no primary user transmission in the current frequency band.

Sensor Node Figure 2a is a block diagram illustrating the main components of a sensor node 5 that operates multiple primary user signal detectors in parallel. As shown, the sensor node 5 includes transceiver circuitry 323 which is operable to transmit signals to and to receive signals from other nodes (e.g. the master node 3) via one or more antennas 325. A controller 327 controls the operation of the transceiver circuitry 323 in accordance with software stored in memory 337. The sensor node also includes a user interface 329 that is controlled by the controller 327 and which allows a user to interact with the sensor device. The software stored in memory 337 includes, among other things, an operating system 339, a communication module 341, a positioning module 343, a sensing manager module 345, a sampling module 347, a plurality of detector modules 349-1 to 349-N, a noise estimation module 351, a decision module 353, a reporting module 355 and a transmission frequency control module 357. In this embodiment, the memory 337 also stores a database 359 of target probabilities and other settings for different types of primary user signals.

The operating system 339 controls the operation of the sensor node 5. The communication module 341 controls communications between the sensor device 5 and external devices via the transceiver circuitry 323 and the antenna 325. The positioning module 343 operates to determine location information for the sensor device. It may be a GPS or similar satellite or terrestrial location determining module. The positioning module 343 sends regular position updates to the master node 3 either on demand or at predetermined time intervals or just together with a reported detection of a primary user transmission. The sensing managing module 345 operates to control the operation of the different detector modules 349 and to control the sampling module 347 so that it provides the required samples to the detector modules 349. The noise estimation module 351 operates to estimate the noise within the frequency band being sensed and to provide this estimate to the detector modules 349 that require noise information to provide a reliable detection.

The detector modules 349 operate, based on information from the database 359 that is selected based on settings defined by the sensing manager module 345, to detect if there is a primary user transmission in a current sensing period. Each detector module 349 uses a different technique to try to perform the detection.

The decision module 353 operates to take a decision on the presence of a primary user transmission within the frequency band being sensed based on the detection results from all of the detector modules 349. The reporting module 355 reports to the master node 3 when the decision module 353 decides that a primary user transmission is present in the monitored frequency band. Finally, the transmission frequency control module 357 operates to receive information transmitted from the master node 3 or another secondary user device, identifying if there is a specific frequency band in which it can transmit opportunistic signals without interfering with a primary user; and to control the transmission frequency used by the transceiver circuitry 323 for such opportunistic transmissions accordingly.

In the above description, the sensor node 5 has been described, for ease of understanding, as having a number of discrete modules (such as the communication module, the sensing manager module, the detector module, the sampling module, the reporting module etc). Whilst these modules may be provided in this way for certain applications, for example where an existing system has been modified to implement the invention, in other applications, for example in systems designed with the inventive features in mind from the outset, these modules may be built into the overall operating system or code and so these modules may not be discernible as discrete entities.

Database In general, when making its decision, the sensor node 5 compares multiple test statistics derived from the signal samples, with respective thresholds. If the received signal contains only noise, then the test statistics should be lower than the respective thresholds; otherwise the test statistics should be greater than the respective thresholds. As shown in Figure 2a, the sensor node 5 has multiple different types of detector modules 349 that each makes the detection based on a different test statistic. For example, an energy detector module estimates the signal energy in the channel and compares it to a threshold and a power detector module estimates the power of the signal in the channel and compares it to a threshold. Performing energy or power detection does not require any knowledge of the signal characteristics. Therefore, energy and power detection are blind detectors. However, a cyclo-stationary detector requires knowledge of the cyclo-stationary features of the signal (which are stored in the database 359). The features are normally non-random spectral or temporal features (such as those related to the symbol period, modulation type etc.) that will be present when the primary signal is present and not present when the primary signal is absent.

Further, while the energy detector and the power detector are sensitive to noise uncertainty, a cyclo-stationary detector is not and therefore the energy detector and the power detector require noise information whilst the cyclo-stationary detectors do not.

The performance of a detector is generally measured by its false alarm probability and its detection probability. The false alarm probability is the probability that the detector decides a primary signal is present when there is actually no primary S signal in the analyzed frequency band. The detection probability is the probability that the detector decides a primary signal is present while a primary signal is actually present in the analyzed frequency band. The main challenge of the sensing is to have the detection probability as high as possible, and the false alarm probability as small as possible. But, in general there is a dilemma: when the detection probability increases, the false alarm probability also increases.

Typically, a target false alarm probability is set for the sensing algorithm and this is used to determine the decision threshold to be used by the detector. At the end of the sensing using the determined decision threshold, the real mean false alarm probability can be determined and this may be different from the target false alarm probability. Therefore, there is a need to configure the detector with a target false alarm probability that will keep the real mean false alarm probability below (but as close as possible to) some defined maximum false alarm detection probability (typically set by a system controller for the secondary users).

In this embodiment, the sensor node 5 stores various data calculated off-line for the different types of primary signals that may be detected and for the different kinds of detector modules 349 that the sensor node 5 has. Figure 2b illustrates the data held within the database 359. As shown, the database 359 maintains a table 369-1 to 369-N, one for each detector modules 349. Each table 369 then stores, for each different type of primary signal (illustrated in the tables as Type 1 to Type M), a sensing duration Ts over which the sensing is to be performed and a target probability of false detection (Target Pf).

The sensing duration Ts depends on the respective individual primary user.

Typically the sensing duration Ts will be between 1 ms and few seconds.

The target probability of false detection (Target Pf) is determined for each detector 349 in advance in order to ensure that the detector's actual probability of false detection is close to but smaller than a desired maximum probability of false detection (Pf,max). This desired maximum probability of false detection depends on the individual primary user (in terms of permitted interference) and on the desired reliability of the secondary user traffic. If it is set at a high value (say 0.5), then interference for the primary user will be low but secondary user communications will be interrupted frequently leading to poor reliability for secondary data communication. If it is set at a low value (say 0.01), then there will be better reliability for secondary user data communication but the primary user is likely to experience more interference compared to the previous case. Typically, Ptmax will be set between 0.01 and 0.2. The way in which the Target Pf is determined will be described in more detail later.

Operation Figure 3 illustrates the functional inter-operation between the main modules forming part of the sensor node 5 shown in Figure 2a. Initially, the sensor node 5 will decide with other nodes (such as with the master node 3) that it wishes to operate in a given frequency band (B) that has been licensed to a particular primary user (such as a television channel). As the type of signal transmitted by the primary user is known, defining the operating frequency band effectively defines the type of signal that the primary user will transmit (e.g. OFDM, QPSK etc.). The sensor node 5 can therefore use the determined operating frequency band to select the data from the column of each table 369 corresponding to the known type of primary signal that is expected in that frequency band. More than one type of primary signal can be expected in a given frequency band (B). For example, the wideband primary user 6 may transmit an OFDM signal, whereas the narrow band primary user 7 may transmit a QPSK signal. The data retrieved from the database 359 identifies, among other things, the sensing duration Ts and the target probability of false alarm (Target Pf) for each of the detector modules 349.

Each of the detector modules 349 uses the corresponding target probability of false alarm to determine a decision threshold for its sensing task. The sensing manager 345 uses the defined sensing duration Ts to control the periods over which the sensor node 5 performs its sensing and reports its sensing results to the master node 3, which makes the final decision about the band occupancy. After each sensing task, with duration Ts, the master node 3 may instruct the sensor node to stop transmitting in a frequency band that has been found to be occupied by the primary user. The sensing manager 345 controls the operation of the sensor node 5 so that it repeats the sensing operation such that at the end of one sensing period a decision can be taken and then a new sensing period begins. In this way, if the primary user starts to transmit within the licensed frequency band (B), the sensor node 5 will be able to stop its own transmissions within the allotted time period associated with that primary user.

Once the sensing manager 345 has setup each of the detectors 349 in accordance with the appropriate data from the database 359, the sensing manager 345 instructs the sampling module 347 to obtain signal samples from the desired operating frequency band (B), which it passes to each of the detector modules 349. The sampling module 347 also forwards samples to the noise estimation module 351, which uses the samples to estimate the noise in the desired operating frequency band (B). In this embodiment, the samples sent to the noise estimation module 351 are from an adjacent (close) frequency band, where the noise level is assumed to be similar to the noise level in the desired operating frequency band (B) and where there should be no primary user transmissions. The noise estimates made by the noise estimation module 351 are then supplied to the detectors 349 that require the noise information to provide an accurate primary signal detection. Features detectors and other noise-insensitive detectors 349 do not need the noise information. At the end of a detection period (ti), the results from each detector module 349 are passed to the decision module 353, which makes a final decision as to whether or not a primary user transmission is detected in the desired operating frequency band from the individual results from the detector modules 349. In particular, in this embodiment, the decision module 353 determines that there is a primary user transmission when at least one detector module 349 detects a primary user transmission. Figure 4 illustrates the parallel operation of the detector modules 349 and the noise estimation module 351 during the overall sensing duration Ts.

Off-line process As discussed above, the database 359 maintains a table of data for each detector 349 that identifies a target probability of false alarm (Target Pf) and a sensing duration Ts. The way in which these entries in the database 359 may be determined will now be explained.

The target false alarm probability is a value, between 0 and 1, used to compute the decision threshold of a detector 349, for a given type of primary signal, such that the actual mean false alarm probability does not exceed the maximum allowed false alarm probability Pf,max. As discussed above, as the detection probability of the detector 349 increases with the actual false alarm probability, the target false alarm probability should be computed such that the actual mean false alarm probability (the mean false alarm probability at the end of the detection process) is just under Pf,max.

For detectors 349 that detect the primary signal by looking at the energy of the signal, the actual mean false alarm probability of the energy detector 349 is a function of the sensing time Is (or the total number of samples within time Ts) and the target false alarm probability only. Therefore, a target false alarm probability for the energy detector 349 can be computed in order to have an actual mean false alarm probability just under Pf,max. The target false alarm probability is computed using example signals in average noise conditions. In the event that the actual signal contains more interference than the average amount used in the off-line process, the actual mean false alarm probability will be less than it was compared to when there is an average amount of noise. Therefore, it is possible to guarantee that the actual false alarm probability will always be lower than Pf,max.

The robustness of feature based detectors 349 (e.g. the cyclo-stationary detector) against noise uncertainty and interference is well-known. The actual mean false alarm probability for such feature based detectors 349 is close to the target false alarm probability. Therefore, the target false alarm probability for the feature based detectors 349 can be set just under Pf,max. In this way, it is possible to guarantee that the actual mean false alarm probability of feature based detectors 349 will be lower than Pf,max.

In addition to, or instead of, direct calculation of the target probability of false detection, an iterative routine such as that shown in Figure 5 may be used to set the target probabilities of false detection for the different detectors 349. As illustrated in Figure 5, initial parameters 401 for a detector 349 are set. These parameters 401 include, for example, the sampling frequency, the number of samples to be processed (which is a function of the sensing duration Ts and the sampling frequency) and the primary signal features such as the type of modulation, the symbol duration, the guard period (for OFDM) etc. Then a candidate target probability of false alarm 403 is chosen from a set 404 of possible values and used to compute the detection threshold used by the detector 349.

Training signals 405 are then applied, some having the primary signal to be detected and typical noise levels and some just having the noise and no primary signal. The detector 349 is then run over the defined number of samples for each training signal and once the training signals have been processed, an actual probability of false alarm (Actual Pf) 407 is computed. A check is then made at 409 to see if the actual probability of false alarm 407 is greater than or equal to the defined maximum probability of false detection (Ff,max). If it is then the target probability of false alarm candidate 403 is discarded and another target probability candidate is selected from the set 404 and the process is repeated.

If the actual probability of false alarm 407 is less than the maximum probability of false alarm Pf,max, then a check is made at 411 to see if the difference between these two probabilities is suitable -i.e. close enough so that in practice the actual probability of false alarm will be close to but less than the defined maximum value (Pf,max); as this will ensure the highest probability of detection for the detector 349. If the actual probability of false alarm 407 is much lower than the defined maximum, then the candidate target probability of false detection 403 is discarded and another one from the set 404 is chosen and the training process repeated.

Once it is determined that the actual probability of false alarm 407 is less than but sufficiently close to the maximum probability of false detection (Pf,max), the current target probability of false detection candidate 403 is set at 413 as the actual target probability of false detection (Target Pf) to be used for the current detector 349 for the defined sensing duration Ts and maximum probability of false alarm (Pf,max) and for the current type of primary signal. The training is then repeated for the next detector 349 until target probabilities of false alarm have been determined for all detectors 349. The determined target probabilities are then used to populate the tables 369 stored in the database 359.

Advantages The embodiment described above offers a number of advantages including: The embodiment benefits from detectors diversity (different detectors having different decisions for a given frequency band) relying on the detectors that are more adequate for the situation thus more likely to detect a primary user when it is really transmitting.

At each sensing time, the false alarm probability of the overall sensing is equal to the false alarm probability of the selected detector. Therefore, the embodiment leads to a lower overall probability of false alarm than prior art systems where sensing methods that combine two different detectors leads to a false alarm probability higher than the false alarm probability of a single detector.

When the set of detectors includes different sensing techniques with specific sensing methods, the proposed sensing scheme allows sensing licensed bands in situations where only one detector is efficient. For example, for the case of the energy detector and the cyclo-stationary detector, the energy detector is known to be unreliable when there is uncertainty in the noise power level and the cyclo-stationary detector could provide reliable sensing result. However, the energy detector provides good sensing results in higher SNR situations and its detection probability can be better than that of the cyclo-stationary detector.

In a fast changing environment where primary users appear and disappear all the time, the embodiment described above can instantaneously choose from a larger set of algorithms, the one that increases the detection performance.

The embodiment provides improved detection performance by instantaneous selection and not by long-term selection (based on long-term performances), which underperforms in fast changing environments.

The embodiment provides a selection algorithm that reacts better to fast changing environments, by instantaneously switching between several detection algorithms, maximizing the long-term detection using short-term observations.

S Modifications and Alternatives Detailed embodiments have been described above. As those skilled in the art will appreciate, a number of modifications and alternatives can be made to the above embodiments whilst still benefiting from the inventions embodied therein. By way of illustration only a number of these alternatives and modifications will now be described.

In the above embodiment, the detectors operated in parallel and output their detection results to a decision module. As those skilled in the art will appreciate, the parallel operation of the detectors may be implemented using multiple processors operating in parallel, but it is more likely to be implemented as multiple processing threads that are processed in parallel on a single processor.

In the above embodiment, one sensing duration (Ts) and one target probability of false alarm (Target Pf) were stored in the database 359 for each primary user signal type. In an alternative embodiment, more than one sensing duration (Ts) and/or more than one target probability of false alarm (Target Pf) can be set for the same primary signal type. In this case, the sensor node 5 would need to be informed of the value of Ts and Target Pf to use at a given point in time for the type of primary signal to expect.

In the embodiment described above, the sensor node 5 operated autonomously with respect to detecting primary user transmissions. In an alternative embodiment, the master node 3 (or indeed any of the other secondary user nodes) may take a more active role in controlling the operation of the sensor node 5. In particular, the database 359 stored in the sensor node 5 may be stored fully or partially in the master node 3 or in another one of the secondary nodes. In this case, the secondary node that maintains the database 359 would inform the sensor node 5 of the desired operating frequency band and would send the sensor node 5 the relevant target probabilities for the different detector modules 349 run by the sensor node 5; together with the sensing duration (Is) and any other information that the sensor node 5 requires to configure its detectors 349.

Figure 6 is a block diagram illustrating the main components of the master node 3 that would be used in such an embodiment. As shown, the master node 3 includes transceiver circuitry 623 which is operable to transmit signals to and to receive signals from other nodes via one or more antennas 625. A controller 627 controls the operation of the transceiver circuitry 623 in accordance with software stored in memory 637. The controller 627 is also able to communicate with other communications devices via a network interface 629. The software stored in memory 637 includes, among other things, an operating system 639, a communication module 641, a positioning module 642, a sensor configuration module 643, a results analysis and decision module 647 and a secondary user control module 649. The master node 3 also includes the above described database 359, which stores the relevant target probabilities of false alarm for the different detector modules in the sensor node for the different primary signal types.

The operating system 639 is operable to control operation of the master node 3.

The communication module 641 provides the functionality to allow the master node to communicate with the sensor nodes 5 via the transceiver circuitry 623 and the antenna 625 and to communicate with other network devices and nodes via the network interface 629 (which may be a copper or optical fiber interface). The positioning module 642 allows the master node 3 to determine the location of the different sensor nodes 5 -either from measurements received from the sensor nodes 5 or by triangulation using signals transmitted to or received from the sensor nodes 5 by a plurality of network nodes (such as the master node 3). The sensor configuration module 643 operates to signal to the sensor node 5 the relevant data from the database 359 to configure the detector modules 349 on the sensor node 5. The sensor node 5 will have to inform the master node in advance what detectors it has, so that the correct configuration data can be sent to the sensor node 5. The results analysis and decision module 647 operates to receive the sensed results back from the sensor node(s) 5 and to make a decision as to whether or not a primary user is currently transmitting within the desired operating frequency band. The secondary user control module 649 operates to inform the secondary users if it possible to transmit opportunistic signals in the desired operating frequency band without interfering with a primary user.

In an embodiment where the master node 3 takes more control, the master node may initiate the spectrum sensing by deciding where, in the overall spectrum, each sensor nodes 5 is to sense. The master node 3 also specifies to the sensor nodes the type of primary signal that they must seek. In response, the sensor nodes 5 can select the relevant data from the column of each table 369 corresponding to the specified type of primary signal (or this information may be supplied by the master node as well). Once the different sensor nodes 5 have reported their sensing results to the master node 3, the master node stores information about spectrum holes that are found (i.e. where there are no detected primary user transmissions) and the master node 3 then decides in which specific frequency band a sensor node 3 may transmit.

In the above embodiments, all of the sensor nodes used to sense for primary user transmissions were secondary users -i.e. they transmitted in the licensed band as well as sensing in the licensed band. In other embodiments, one or more of the sensor nodes may not be secondary users. Instead, they may simply sense within the licensed band and report their sensing results to the master node or to the other secondary users.

In the above embodiments, the licensed band corresponded to a DVB-T channel of 6MHz or 8MHz bandwidth. As those skilled in the art will appreciate, there are other potential DVB-T configurations, and the standardization includes a 5MHz and 7MHz channel option. Additionally, whilst the narrow band primary transmitter operated within a bandwidth of about 200kHz, devices operating in other bandwidths can be used. For example, an analogue PMSE device uses FM modulation and the bandwidth is about 200kHz, however, a digital PMSE device (which use digital modulation such as QASK) can have a bandwidth of 400kHz or 600kHz. Additionally, the operating band of the secondary users may be greater than, and overlap with, the whole or part of the licensed band (B).

In an embodiment where the invention is implemented in an LTE system (as defined by the 3GPP standards), the master node will typically be formed by an LTE base station. Multiple base stations are provided in an LTE system and they may each operate as master nodes controlling sensor nodes (typically user devices such as cellular telephones) within its cell(s). The base stations may operate autonomously in a de-centralised manner taking decisions about their respective localities without communicating with a higher network entity. The base stations may also co-operate with each other exchanging information over their "X2" interface and taking decisions based on information received from neighbouring base stations.

In the above embodiment, a number of nodes have been described. As those skilled in the art will appreciate, such nodes may comprise any kind of communications node or device, including access points and user devices such as, for example, mobile telephones, personal digital assistants, laptop computers, web browsers, etc. In the above embodiments, a number of software modules were described. As those skilled in the art will appreciate, the software modules may be provided in compiled or un-compiled form and may be supplied to the node as a signal over a computer network, or on a recording medium. Further, the functionality performed by part or all of this software may be performed using one or more dedicated hardware circuits. However, the use of software modules is preferred as it facilitates the updating of the node in order to update its functionality. Similarly, although the above embodiments employed transceiver circuitry, at least some of the functionality of the transceiver circuitry can be performed by software.

Various other modifications will be apparent to those skilled in the art and will not be described in further detail here.

Claims (1)

1. <claim-text>Claims 1. A communication node for use in a cognitive radio communications system, the communication node comprising: transceiver circuitry for transmitting signals to and for receiving signals from at least one other communication node within a desired operating frequency band; a plurality of detectors being configured to operate in parallel to process signals received by the transceiver circuitry using a respective different processing technique to determine if a primary user of the desired operating frequency band is currently transmitting within the desired operating frequency band; and a communications controller operable to control the transmissions of the communication node in dependence upon detection results made by said plurality of detectors.</claim-text> <claim-text>2. A communication node according to claim 1, comprising a decision module operable to receive the detection results from all of said detectors and operable to decide whether or not the primary user of the desired operating band is currently transmitting within that frequency band.</claim-text> <claim-text>3. A communication node according to claim 2, wherein said decision module is operable to decide to stop transmitting within the desired operating frequency band if any of the detectors detects the primary user transmissions.</claim-text> <claim-text>4. A communication node according to claim 2 or 3, wherein said decision module is operable to decide to allow the communication node to transmit within the desired operating frequency band if all of the detectors do not detect primary user transmissions in the desired operating frequency band.</claim-text> <claim-text>5. A communication node according to any of claims 2 to 4, wherein said communications controller is operable to control the transmissions of the communications node in dependence upon the decision of the decision module.</claim-text> <claim-text>6. A communication node according to any of claims 1 to 5, wherein said detectors are configurable to detect for different types of primary user transmissions and comprising a sensing manager operable to configure the detector modules in accordance with configuration data that depends upon an expected type of primary user transmission for the desired operating frequency band.</claim-text> <claim-text>7. A communication node according to claim 6, wherein the configuration data includes data relating to a respective detection threshold for each of the detectors.</claim-text> <claim-text>8. A communication node according to claim 7, wherein the data relating to a detection threshold for a detector defines a target probability of false detection and wherein each detector is operable to determine a detection threshold in dependence upon the defined target probability of false detection.</claim-text> <claim-text>9. A communication node according to claim 7 or 8, wherein the configuration data for a detector includes data defining a sensing duration over which the detector is to sense for the primary user transmissions.</claim-text> <claim-text>10. A communication node according to any of claims 6 to 9, comprising a database of configuration data for different types of primary user transmissions and for the different detectors.</claim-text> <claim-text>11. A communication node according to claim 10, operable to receive an indication of a primary user signal type from a remote device and operable to retrieve configuration data from said database corresponding to the indicated type of primary user signal.</claim-text> <claim-text>12. A communication node according to any of claims 1 to 11, further comprising a noise estimator operable to estimate a noise level within the desired operating frequency band and wherein a subset of the detector modules are operable to use the estimated noise level when processing the signals received by the transceiver circuitry to determine if the primary user of the desired operating frequency band is currently transmitting within the desired operating frequency band.</claim-text> <claim-text>13. A communication node according to any of claims 1 to 12, wherein one or more of said detectors are operable to determine the energy of signals within the desired operating frequency band and wherein one or more other detectors are operable to detect at least one cyclo-stationary feature within the received signal that indicates the presence of the primary user transmission.</claim-text> <claim-text>14. A communication node for use in a cognitive radio communications system, the communication node comprising: transceiver circuitry for transmitting signals to and for receiving signals from at least one other communication node; a database storing configuration data for a plurality of detectors that are used by the at least one other communication node, the configuration data including, for each detector, data relating to a detection threshold to be used by the detector for each of a plurality of different types of primary user signals; and a configuration module operable to send configuration data for the detectors in dependence upon a type of primary user signal that is expected to be transmitted within a desired operating frequency band.</claim-text> <claim-text>15. A communication node according to claim 14, wherein the data relating to a detection threshold for a detector defines a target probability of false detection.</claim-text> <claim-text>16. A communication node according to claim 15 or 16, wherein the configuration data for a detector includes data defining a sensing duration over which the detector is to sense for the primary user transmissions.</claim-text> <claim-text>17. A method performed by a communication node in a cognitive radio communications system, the method comprising: transmitting signals to and receiving signals from at least one other communication node within a desired operating frequency band; using a plurality of detectors that operate in parallel to process signals received by the transceiver circuitry using a respective different processing technique to determine if a primary user of the desired operating frequency band is currently transmitting within the desired operating frequency band; and controlling the transmissions of the communication node in dependence upon detection results made by said plurality of detectors.</claim-text> <claim-text>18. A method performed by a communication node in a cognitive radio communications system, the method comprising: transmitting signals to and receiving signals from at least one other communication node; storing a database having configuration data for a plurality of detectors that are used by the at least one other communication node, the configuration data including, for each detector, data relating to a detection threshold to be used by the detector for each of a plurality of different types of primary user signals; and transmitting configuration data for the detectors in dependence upon a type of primary user signal that is expected to be transmitted within a desired operating frequency band.</claim-text> <claim-text>19. A computer implementable product comprising computer implementable instructions for causing a programmable computer device to be configured as the communication node of any of claims 1 to 16.</claim-text>