Distance Information Based Relay Selection Scheme For IR-UWB Networks

With the development of various wireless technologies, the demand for mobile communications are increasing. The next generation of wireless communications requires realizing seamless worldwide coverage, at the same time, can implement high-speed multimedia communication including voice, text, images and video and low-speed perception business including detection, positioning, and imaging. IR-UWB technology, with its low power consumption, low cost, high transmission speed, anti-jamming and other unique advantages aroused widespread concern in the 802.15.4a standard low-speed communications. It's also the main research object in other areas. However, UWB signals suffer from harsh radio channel environment like other wireless transmission technologies. And FCC set strict power constraints for UWB signals. As a result, the performance and coverage of UWB systems are limited seriously. Cooperative systems provide spatial diversity gain without requiring the establishment of multiple antennas on the source or destination nodes, and thereby boost the transmission performance significantly. This is especially attractive for UWB ad hoc networks where terminals are low cost, low power and extremely small in size. Among the variety of cooperative transmission protocols, opportunistic relaying is more attractive than others due to the avoidance of coordination between the relay nodes and its superior performance. The key problem in opportunistic relaying is the way to choose the best relay out of all available relays. This thesis focuses on partner selection algorithms and explores the combination of cooperative communication technology with ultra-wideband networks to expand the coverage area and improve system performance in the case of satisfying the strict power constraint.The main contributions of this thesis can be summarized as follows:1.Making use of the ranging capability of UWB systems, this paper presents a distance-information-based opportunistic relay selection algorithm. In this algorithm, taking the distances between the source, idle neighbors and the destination as (dsi,did), the proposed algorithm selects one of the neighbor nodes which is near to source and destination as the best relay by using the min-max optimization principle. The proposed algorithm doesn't need to estimate channel gains, but just need to know the relative distance information of relay node. So the proposed algorithm is easy to implement in comparison with the classic ORS algorithm. In comparison with random relay selection algorithm, the proposed algorithm is more reliable.2. The performances of proposed algorithm is evaluated and compared with that of the classic ORS algorithm and random relay selection algorithm by simulations in narrowband and IR-UWB systems. The simulation results show that:1) Compared with no collaborative communications system, remarkable diversity gains can be obtained by using the proposed relay selection algorithm in both narrowband and IR-UWB systems. In the narrowband system, the diversity gain is about 2 dB when the bit error rate is 10-4.While in IR-UWB system the diversity gain is more significant, say more than 5 dB.2) In both narrowband and IR-UWB systems, the proposed distance-information-based opportunity relay selection algorithm achieves almost the same diversity gains as that of the classic ORS algorithm. It is worth noting that the proposed algorithm lower complexity. And the performance of all kinds of algorithms is improved s as alternative relay node number increase.3. In IR-UWB system, the influence of ranging error to relay selection and the performance of system is investigated. When the ranging error variance is 10-3 m, the error rate of the best relay selection is nearly zero. And the error rate of the best relay node selection increased with the increased variance. When the ranging error variance is 0.025, the diversity gains when the bit error rate is 10-5 is just less 0.3dB than that of the situation with no ranging error. When the ranging error variance is 10m, the diversity gains is also more nearly 1dB than that of the random relay selection scheme when the bit error rate is 10-5.