Research On Performance Of Diversity Receive And Relay Systems In Double-Rayleigh Fading

To provide high-quality and high-speed information transfer services has always been thepursuit of wireless communication developments. In order to approach this goal under particularfading distributions of channel transmission coefficients, the capacitiy of wireless communica-tion system should be estimated, or the parameters of the outage probability (OC), the averagesymbol error rate (ASER) or the average bit error rate (ABER) on the system should be calcula-ted in the fulfillments of certain transmission rate requirements. Hence the transmission rate andreliability of the communication system can be evaluated accurately.Along with the technological developments in wireless ad hoc networks, vehicular ad hocnetworks and cellular networks, a small-scale fading model named as cascade Rayleigh ordouble-Rayleigh attracts extensive attentions. Experimental measurements and theoreticalanalysis both indicate that the wireless channels are subject to double-Rayleigh fadingdistribution in some scenarios; outdoor-to-indoor in specific propagation environments, vehicle-to-vehicle under the conditions of non-line-of-sight propagation, and mobile-to-mobilepropagation through a keyhole.Diversity receive is one of the traditional techniques to combat the effects of fading. As theterminals share their antennas and other resources to create a “virtual array” through distributedtransmission and signal processing, the newly-emerging cooperative relay technology canmitigate the influences of channel fading. On the basis of the conclusions on the performance ofa single-input single-output (SISO) system in double-Rayleigh fading, the thesis focuses on thediversity receive system employing different combining techniques and the cooperative relaysystem with a full or partial receiving-channel-state-information (RCSI)-assisted relay overdouble-Rayleigh fading channels. It gives them more comprehensive and in-depth performanceinvestigations in terms of OCs, ASERs and diversity gains. The main research work issummarized as follows:1. The performance on a SISO system is studied over a double-Rayleigh faded channel. Theclosed-form expression on high order moments is provided for a double-Rayleigh distributedrandom variable. The high order moments of the output signal-to-noise ratio (SNR) are obtainedand the exact solution to OC is concluded for the SISO system in double-Rayleigh fading. Themoment generating function (MGF) of the output SNR is derived for the system. A generalizedexpression on ASER for performance evaluation of various M-ary modulations is acquiredthrough MGF-based approach, and the ASER formulas of binary modulation systems are givencovering coherent and non-coherent demodulation schemes.2. The performance of a diversity receive system is explored under double-Rayleigh fading conditions. The closed-form solutions to high order moments of output SNRs are presented formulti-branch maximal ratio combining (MRC) and equal gain combining (EGC) receivers and adual-branch selection combining (SC) receiver. The MGF of the output SNR is acquired for theMRC receive system with diversity branches operating in unbalanced double-Rayleigh fadingamplitudes, hence its ASER and diversity performances are further discussed through MGF-based approach. The approximate rational expressions on the MGF of the output SNR of an EGCreceiver are obtained by using Padé approximation (PA) technique and based on them the appro-ximate solutions to OC and ASER are presented for the system. The closed-form expression onOC is derived for a multi-branch SC system. The MGF of the output SNR of a dual-branch SCsystem is obtained and the solutions to its ASER are concluded for binary modulations.3. Under the assumptions that the relay fixed-positioned above roof-top level or at streetlevel has access to full RCSI (FRCSI) or partial RCSI (PRCSI), the error performance on thefixed-node relay system employing amplify-and-forward (AF) relaying mode is investigated inthe case of Rayleigh faded source-to-relay, relay-to-destination links and a double-Rayleighfaded source-to-destination link. The optimum power allocation (OPA) methods are presentedcovering the scheme maximizing the instantaneous output SNR of the destination node (OPA-I)and the one minimizing the ASER of the system (OPA-II). The closed-form solution to the powerallocation coefficient (PAC) of OPA-I is derived assuming FRCSI fixed-node relaying. Thenumerical solutions to PACs of OPA-II are calculated for FRCSI and PRCSI fixed-node relaysystems in combination with three specific topological structures. The exact solution to OC andthe closed-form expression on MGF of the output SNR are concluded for the FRCSI fixed-noderelay system, and the upper bound on ASER is provided. For the PRCSI fixed-node relay system,the approximate expressions on the MGF of the output SNR are derived with the use of PAmethod, and from an alternative expression of Gaussian Q-function the upper bounds onpairwise error probability (PEP) together with ASER are deduced.4．Under the assumptions that the relay being a mobile node of the same properties as thesource or destination node has access to FRCSI or PRCSI, the outage and ASER performances ofa mobile node relay system employing AF relaying mode are investigated over double-Rayleighfaded source-to-relay, relay-to-destination and source-to-destination links. The numericalsolutions to PACs of OPA-II are calculated for PRCSI mobile node relay systems in combinationwith three specific topological structures. The lower bounds on OC and ASER of the FRCSImobile node relay system are derived by using the inequality between geometric and harmonicmeans. The upper bounds on PEP and ASER are presented for both FRCSI and PRCSI mobilenode relay systems.Theoretical analysis and simulation results reveal a lot of useful conclusions, mainly including:1. As far as the double-Rayleigh faded SISO system is concerned, the fluctuations of theinstantaneous output SNR around its mean are more severe than in Rayleigh fading, thereby itsoutage probability and ASER perform worse. The system can achieve full diversity order of uni-ty as the average SNR tends to infinity but cannot get it within practical receiving SNR ranges.2. Due to the fact that the instantaneous output SNR fluctuates around its mean severely indouble-Rayleigh fading than in Rayleigh fading, the outage or ASER performance of thediversity receive system becomes worse accordingly in double-Rayleigh fading as far as thesame combing technique is concerned. In the comparisons of MRC, EGC and SC each other indouble-Rayleigh fading, MRC results in best performance whereas SC results in worst perfor-mance, but the performance difference between SC and MRC is smaller than that in Rayleighfading. The diversity gains achievable of the diversity receive system in double-Rayleigh fadinggo up slower with the increase in SNR than in Rayleigh fading, hence there exists an apparentgap between them and full diversity gains within practical receiving SNR ranges.3. Compared with the fixed-node relay system where only source-to-relay link is subject todouble-Rayleigh fading distribution, the mobile node relay system achieves fewer diversity gainsat given SNR owing to three underlying double-Rayleigh faded transmission links. For thePRCSI relay system either the relay being a fixed node or a mobile node, the diversity gainsachievable approach to their full gains of two with the increase in SNR slower than in FRCSIrelaying scenario due to the relay not having access to accurate RCSI.4. Either a fixed node or a mobile node acts as a relay, the corresponding system employingOPA-II method behaves worse in capacity and error performance than that employing OPA-Imethod whereas it approximates to that employing equal power allocation (EPA) method as therelay approaches to the source, but the performances of the two OPA methods are equivalent toeach other while the relay is close to the destination and therefore both of them are in gains ofabout3dB in SNR in comparision to EPA solution. For the case that the relay approaches to thesource and OPA-II method is employed, the PAC of the PRCSI fixed node relay system appro-ximates to0.5-this is just similar to a delayed-transmission diversity system of two collocatedantennas, the PAC of the FRCSI fixed node relay system is slightly smaller than0.5, and thePAC of the PRCSI mobile node relay system calculated by using different path loss coefficientranges from0.55to0.7.