Software Defined Radio Solutions for Wireless Communications Systems
|Kustantaja||Tampere University of Technology|
|Tila||Julkaistu - 23 marraskuuta 2018|
|Nimi||Tampere University of Technology. Publication|
This thesis mainly addresses the challenges related to SDR-based implementation of today’s wireless devices. One of the main targets of most of the wireless standards has been to improve the achievable data rates, which imposes strict requirements on the processing platforms. Realizing real-time processing of high throughput signal processing algorithms using SDR-based platforms while maintaining energy consumption close to conventional approaches is a challenging topic that is addressed in this thesis.
Firstly, this thesis concentrates on the challenges of a real-time software-based implementation for the very high throughput (VHT) Institute of Electrical and Electronics Engineers (IEEE) 802.11ac amendment from the wireless local area networks (WLAN) family, where an SDR-based solution is introduced for the frequency-domain baseband processing of a multiple-input multipleoutput (MIMO) transmitter and receiver. The feasibility of the implementation is evaluated with respect to the number of clock cycles and the consumed power. Furthermore, a digital front-end (DFE) concept is developed for the IEEE 802.11ac receiver, where the 80 MHz waveform is divided to two 40 MHz signals. This is carried out through time-domain digital ﬁltering and decimation, which is challenging due to the latency and cyclic preﬁx (CP) budget of the receiver. Different multi-rate channelization architectures are developed, and the software implementation is presented and evaluated in terms of execution time, number of clock cycles, power, and energy consumption on different multi-core platforms.
Secondly, this thesis addresses selected advanced techniques developed to realize inband fullduplex (IBFD) systems, which aim at improving spectral efﬁciency in today’s congested radio spectrum. IBFD refers to concurrent transmission and reception on the same frequency band, where the main challenge to combat is the strong self-interference (SI). In this thesis, an SDRbased solution is introduced, which is capable of real-time mitigation of the SI signal. The implementation results show possibility of achieving real-time sufﬁcient SI suppression under time-varying environments using low-power, mobile-scale multi-core processing platforms. To investigate the challenges associated with SDR implementations for mobile-scale devices with limited processing and power resources, processing platforms suitable for hand-held devices are selected in this thesis work. On the baseband processing side, a very long instruction word (VLIW) processor, optimized for wireless communication applications, is utilized. Furthermore, in the solutions presented for the DFE processing and the digital SI canceller, commercial off-the-shelf (COTS) multi-core central processing units (CPUs) and graphics processing units (GPUs) are used with the aim of investigating the performance enhancement achieved by utilizing parallel processing.
Overall, this thesis provides solutions to the challenges of low-power, and real-time software-based implementation of computationally intensive signal processing algorithms for the current and future communications systems.