Final Report Summary - CUWLS (Coexistence of Ultra-Wideband and Licensed Systems)
Coexistence is one of the crucial design issues in wireless communications. For peaceful coexistence of existing systems and newly deployed technologies, coexistence mechanisms have been studied and developed in both academia and industry. In the standardization of new wireless technologies within the IEEE 802 Local Area Network and Metropolitan Area Network standards, the coexistence issue is considered by the IEEE 802.19 Wireless Coexistence Technical Advisory Group. The coexistence issue is one of the main standardization issues as it is not desired to have the emerging communication system cause interference to existing systems and vice versa.
As the spectrum gets crowded with many licensed systems, there has been a need to consider alternative communication systems. Ultra-wideband (UWB) communications is one of them. In UWB communications, due to very low power spectral density (PSD) of the transmitted signal, the UWB system can operate as an unlicensed underlay system in the same frequency bands of licensed systems. Despite the low PSD of such systems, the European and Japanese regulatory agencies have made the implementation of detect-and-avoid (DAA) techniques mandatory in some common bands so as to protect the usage rights of licensed users. Therefore, the challenge is to implement UWB systems that can coexist with licensed communication systems peacefully. Accordingly, this project investigates the implementation of UWB systems for peaceful coexistence with licensed systems considering the following issues:
(i) “Detection” of licensed systems in order to decide whether the common band is in use or not.
(ii) “Avoidance” of licensed systems if the licensed system is active most of the time.
This project has two main objectives, one for the detection part and one for the avoidance part. Firstly, the project considers licensed system operation in multiple bands, where these systems are operating “dependently”. Hence, the first main objective is to investigate interdependencies of multiple bands so as to improve the primary user detection performance. Secondly, the project considers the linear combination pulses method as suggested by the IEEE 802.15.4a standard for peaceful coexistence. Accordingly, the second main objective is to investigate the feasibility and performance of possible pulse shapes that can be implemented based on the IEEE 802.15.4a standard. In order to reach the two main objectives, the research conducted can be summarized as follows.
For the “detection” part, initially the “uplink-downlink interdependency” in a frequency division duplex system has been investigated. Assuming the uplink-downlink traffic information is available, the bands were jointly processed using the maximum a posteriori based approach and compared with the independent processing of each band using energy detection, which is optimized with the Neyman-Pearson test based approach. It was shown that the detection gain of joint detection over independent detection depended on the bands’ joint activity values and signal-to-noise-ratio (SNR) regions. As an extension of this work, the detection of M independent primary users with known traffic information was considered. Accordingly, the detection in 2-bands (i.e. uplink-downlink) was generalized to M-bands. This is an important case as the UWB system may be sharing the frequency band with multiple systems, and it has to detect all of them before it can communicate. For this generalized case, in addition to quantifying the gain of joint detection over independent detection in terms of probabilities of false alarm and detection for practical scenarios, the operation capability of secondary systems were presented in terms of the fractions of time the secondary system can access the channel without interfering with the primary systems. One common assumption in uplink-downlink detection and M-band detection was that a single UWB system performed spectrum sensing. However, in practice there are multiple UWB systems in a wireless sensor network (WSN), where each system assesses the spectrum individually and sends its decision to the fusion centre for an improved detection performance. Accordingly, an IEEE 802.15.4a based WSN is investigated next, where the detection of an uplink-downlink primary system with multiple UWB systems is considered. IEEE 802.15.4a standard specific modulation formats have been implemented for multiple UWB users in an IEEE 802.15.4a based WSN and the detection performances have been investigated. It was shown that the two standard-specific modulation formats (i.e. coherent and noncoherent modulations) may perform better or worse compared to each other depending on the probabilities of false alarm and miss-detection at each UWB sensor. The results of the detection part are important as the detection performance of UWB systems can be improved significantly by considering available interdependencies in orthogonal frequency bands and by implementing UWB WSNs.
For the “avoidance” part, initially possible pulse shapes that generate notches at desired frequencies were investigated and it was shown that the designed pulses could only generate one or two notches due to the use of limited number of coefficients. Therefore, the focus of the next step was to modify the IEEE 802.15.4a based UWB transceiver structures, which use the designed pulse shapes and can support two different modulation formats. In the presence of a narrowband primary user, it was shown that the modified UWB transceiver structure that uses the IEEE 802.15.4a standard compliant pulse shape can achieve almost the same performance (only 1-2 dB inferior) as the no-interference case. As an extension of this work, the effects of pulse design and interference types were investigated. The primary systems were modeled as orthogonal frequency-division multiplexing (OFDM) based wideband systems, where the effects of bandwidths and number of subcarriers were investigated for various pulses designed. It was shown that employing higher order linearly combined pulses compensated better for the system performance due to having wider notches and more efficient spectrum shaping in the presence of wideband interference. Furthermore, the bit-error rate performance of the UWB system was shown to improve with increasing bandwidth of the OFDM system, which can be explained by the per-subcarrier interference power spectral density decreasing as the bandwidth increasing for constant power interference. In addition, it was observed that the UWB system showed better performance in the presence of an OFDM system with large number of subcarriers in the low signal-to-interference ratio (SIR) region and better performance for small number of subcarriers in the high SIR region. The results of the avoidance part are important as the modified transceiver structure can achieve a reasonable system performance while complying with the European and Japanese regulatory agency mandates.
The overall project results are important in achieving peaceful coexistence between licensed and UWB systems, while improving the detection performance and bit-error rate performance of UWB systems. The potential impact of these results can be two-fold. From the detection point of view, the research results obtained in this project are of interest to both academic research community and industrial research community. While the research results were obtained for UWB systems, these results are directly applicable to cognitive radios, which perform spectrum sensing. Hence, interdependency of multiband systems investigated in this project can be used in improving the primary user detection performance in UWB systems and future cognitive radio systems, if the traffic information of the primary systems is available. From the avoidance point of view, the research results obtained in this project are of interest to the industrial community that will design IEEE 802.15.4a based commercial products. The pulse shapes and modified UWB transceiver structures proposed in this project can be used in prototypes, and their systems performances can be predicted in advance in the presence of narrowband and wideband primary systems considering the UWB system performances reported in the project. All in all, the project results are expected to have a positive impact on future wireless communications technology.
As the spectrum gets crowded with many licensed systems, there has been a need to consider alternative communication systems. Ultra-wideband (UWB) communications is one of them. In UWB communications, due to very low power spectral density (PSD) of the transmitted signal, the UWB system can operate as an unlicensed underlay system in the same frequency bands of licensed systems. Despite the low PSD of such systems, the European and Japanese regulatory agencies have made the implementation of detect-and-avoid (DAA) techniques mandatory in some common bands so as to protect the usage rights of licensed users. Therefore, the challenge is to implement UWB systems that can coexist with licensed communication systems peacefully. Accordingly, this project investigates the implementation of UWB systems for peaceful coexistence with licensed systems considering the following issues:
(i) “Detection” of licensed systems in order to decide whether the common band is in use or not.
(ii) “Avoidance” of licensed systems if the licensed system is active most of the time.
This project has two main objectives, one for the detection part and one for the avoidance part. Firstly, the project considers licensed system operation in multiple bands, where these systems are operating “dependently”. Hence, the first main objective is to investigate interdependencies of multiple bands so as to improve the primary user detection performance. Secondly, the project considers the linear combination pulses method as suggested by the IEEE 802.15.4a standard for peaceful coexistence. Accordingly, the second main objective is to investigate the feasibility and performance of possible pulse shapes that can be implemented based on the IEEE 802.15.4a standard. In order to reach the two main objectives, the research conducted can be summarized as follows.
For the “detection” part, initially the “uplink-downlink interdependency” in a frequency division duplex system has been investigated. Assuming the uplink-downlink traffic information is available, the bands were jointly processed using the maximum a posteriori based approach and compared with the independent processing of each band using energy detection, which is optimized with the Neyman-Pearson test based approach. It was shown that the detection gain of joint detection over independent detection depended on the bands’ joint activity values and signal-to-noise-ratio (SNR) regions. As an extension of this work, the detection of M independent primary users with known traffic information was considered. Accordingly, the detection in 2-bands (i.e. uplink-downlink) was generalized to M-bands. This is an important case as the UWB system may be sharing the frequency band with multiple systems, and it has to detect all of them before it can communicate. For this generalized case, in addition to quantifying the gain of joint detection over independent detection in terms of probabilities of false alarm and detection for practical scenarios, the operation capability of secondary systems were presented in terms of the fractions of time the secondary system can access the channel without interfering with the primary systems. One common assumption in uplink-downlink detection and M-band detection was that a single UWB system performed spectrum sensing. However, in practice there are multiple UWB systems in a wireless sensor network (WSN), where each system assesses the spectrum individually and sends its decision to the fusion centre for an improved detection performance. Accordingly, an IEEE 802.15.4a based WSN is investigated next, where the detection of an uplink-downlink primary system with multiple UWB systems is considered. IEEE 802.15.4a standard specific modulation formats have been implemented for multiple UWB users in an IEEE 802.15.4a based WSN and the detection performances have been investigated. It was shown that the two standard-specific modulation formats (i.e. coherent and noncoherent modulations) may perform better or worse compared to each other depending on the probabilities of false alarm and miss-detection at each UWB sensor. The results of the detection part are important as the detection performance of UWB systems can be improved significantly by considering available interdependencies in orthogonal frequency bands and by implementing UWB WSNs.
For the “avoidance” part, initially possible pulse shapes that generate notches at desired frequencies were investigated and it was shown that the designed pulses could only generate one or two notches due to the use of limited number of coefficients. Therefore, the focus of the next step was to modify the IEEE 802.15.4a based UWB transceiver structures, which use the designed pulse shapes and can support two different modulation formats. In the presence of a narrowband primary user, it was shown that the modified UWB transceiver structure that uses the IEEE 802.15.4a standard compliant pulse shape can achieve almost the same performance (only 1-2 dB inferior) as the no-interference case. As an extension of this work, the effects of pulse design and interference types were investigated. The primary systems were modeled as orthogonal frequency-division multiplexing (OFDM) based wideband systems, where the effects of bandwidths and number of subcarriers were investigated for various pulses designed. It was shown that employing higher order linearly combined pulses compensated better for the system performance due to having wider notches and more efficient spectrum shaping in the presence of wideband interference. Furthermore, the bit-error rate performance of the UWB system was shown to improve with increasing bandwidth of the OFDM system, which can be explained by the per-subcarrier interference power spectral density decreasing as the bandwidth increasing for constant power interference. In addition, it was observed that the UWB system showed better performance in the presence of an OFDM system with large number of subcarriers in the low signal-to-interference ratio (SIR) region and better performance for small number of subcarriers in the high SIR region. The results of the avoidance part are important as the modified transceiver structure can achieve a reasonable system performance while complying with the European and Japanese regulatory agency mandates.
The overall project results are important in achieving peaceful coexistence between licensed and UWB systems, while improving the detection performance and bit-error rate performance of UWB systems. The potential impact of these results can be two-fold. From the detection point of view, the research results obtained in this project are of interest to both academic research community and industrial research community. While the research results were obtained for UWB systems, these results are directly applicable to cognitive radios, which perform spectrum sensing. Hence, interdependency of multiband systems investigated in this project can be used in improving the primary user detection performance in UWB systems and future cognitive radio systems, if the traffic information of the primary systems is available. From the avoidance point of view, the research results obtained in this project are of interest to the industrial community that will design IEEE 802.15.4a based commercial products. The pulse shapes and modified UWB transceiver structures proposed in this project can be used in prototypes, and their systems performances can be predicted in advance in the presence of narrowband and wideband primary systems considering the UWB system performances reported in the project. All in all, the project results are expected to have a positive impact on future wireless communications technology.