Cooperative Transmission In Wireless Communication Systems

With the growing demand for high data rates and high quality of service,there iscontineous evolution in wireless network architecture and advanced transmissiontechnique.It is well known that multiple-input multiple-output (MIMO) technologypromises significant improvements in terms of spectral efficiency and link reliability,and is adopted in many standards.By employing multiple antennas at the transmitterand the receiver,spatial diversity can be achieved with simple space-time blockcoding.However,due to size,cost and hardware complexity limitations,the use ofmultiple antennas at mobile terminals might be impractical in many instances.Recently,cooperative communication has been proposed as a promising technologyfor future wireless communication systems.By exploiting the broadcast nature of thewireless medium,multiple wireless terminals work together to form a virtual antennaarray and spatial diversity is achievable even for single-antenna terminals.Thebenefits of cooperative communication include increasing the spatial diversity order,reducing the transmit power and extending the radio cover range.This noveltransmission technique has attracted a lot of research attention in recent years,and is apromising technique for future LTE-Advanced systems.So far,several efficient cooperative transmission protocols have been proposed inthe literature,including amplify-and-forward (AF),decode-and-forward (DF),andcompress-and-forward (CF).In this thesis,we mainly focus on performance analysisand optimization of amplify-and-forward protocol,cooperative partner selection inmulti-user systems,and novel cooperation protocol design.In an amplify-and-forward cooperative system,the end-to-end signal-to-noiseratio (SNR) is in the form of the harmonic mean of the source-relay channel SNR andthe relay-destination channel SNR.It is quite difficult to derive the exact symbol errorrate (SER) expression.Instead of deriving the exact probability density function (PDF)of the harmonic mean of two independent non-negative random variables,we studythe series expansion of this PDF around zero.This statistical result is then applied toevaluate the end-to-end performance of various AF cooperative systems,includingdual-hop orthogonal AF system,dual-hop selection relaying AF system and multihopAF system.Closed-form expressions for the average SER in the high SNR regime arederived.The presented method is suitable for arbitrary fading channel,provided thatthe PDF of the channel SNR has a valid Taylor expression in the neighborhood of zero.Based on these SER expressions,we further investigate the power allocation indual-hop orthogonal AF systems over Nakagami-m fading channels.Based on thestatistical channel knowledge,the optimal power allocation scheme is proposedaiming at minimizing the asymptotic SER and the outage probability.Since theoptimal power allocation scheme requires solving nonlinear equations,we alsopropose a low complexity sub-optimal power allocation scheme for practicalimplementation.The performance improvement with optimal and sub-optimal powerallocation is analyzed and validated by numeric results.Distributed space-time coding (DSTC) is an efficient cooperation protocol forsystem with multiple cooperative terminals (relays).However,it requires accuratesymbol-level synchronization and prior coordination between cooperative relayterminals,which is difficult to implement in distributed networks.In this thesis,wepropose a novel cooperative transmission scheme based on interleaving to achievecooperative diversity in both synchronous and asynchronous systems with littleprotocol overhead.A low complexity iterative detection algorithm is also proposed tocombine signals from different relays at the receiver.In the proposed scheme,eachcooperative terminal simply interleaves the detected bits using a pre-allocatedinterleaver.The proposed scheme supports arbitrary number of cooperative terminalswithout data rate loss while full-rate full-diversity orthogonal space-time codes don'texist in most cases.The cooperative spatial diversity can be achieved by a lowcomplexity symbol-by-symbol iterative detection algorithm at the receiver even if thereceived signal is asynchronous.So far,most of previous works assume that perfect channel state information (CSI)of all links is available at the destination.In a practical scenario,these CSIs areunknown and must be estimated by the receiver for coherent detection.Comparedwith point-to-point systems,channel estimation in distributed wireless networksbecomes more complicate.Furthermore,when amplify-and-forward relaying protocolis employed,the relay nodes forward the useful signal as well as the noise to thedestination at the same time,which makes the channel estimation problem morechallenging.In this thesis,we propose a novel differential modulation scheme forwireless relay networks with amplify-and-forward relaying,which can bridge the 3dBperformance gap between conventional differential modulation and coherent detection.In the proposed scheme,each transmitted frame is divided into several blocks.Thefirst symbol in each block is referred as reference symbol,and the other symbols in each block are normal symbols.We also investigate the power allocation among thereference symbol and normal symbols to enhance the system performance.Simulationresults are presented to validate the proposed scheme.It is shown that the optimizedsystem performance is very close to coherent receiver in slow fading channels.In a multiuser cooperative communication system,one basic question is"whohelps whom",i.e.how to appropriately match users for cooperative transmission.Inthis thesis,we propose a channel-adaptive partner selection scheme with real-timefairness constrain.During each cooperative transmission,the partners are selectedaccording to the direct links' quality.The user with the worst direct link has thehighest priority to select its partner,while the user with the best direct link is the lastone to select its partner.Also,each user can be selected only once to ensure fairness.By doing so,the proposed partner selection scheme provides a good trade-off betweenfairness and system performance.The lower bound of the BER performance of theproposed scheme is also presented.Selection relaying is a promising technique for practical implementation ofcooperative systems with multiple relay nodes.However,to select the best relay,global channel knowledge is required at the selecting entity,which may result inconsiderable signaling overhead.In this thesis,we consider the relay selectionproblem in dual-hop amplify-and-forward communication systems with partial CSI.We present relay selection strategies aiming at minimizing either the outageprobability or the bit error rate (BER) with different kinds of channel knowledgeavailable,including perfect instantaneous,quantized and statistical CSI.In case ofwhen quantized CSI is available,we propose a target rate based quantizer toefficiently partition the SNR range for outage minimized relay selection,and a targetBER based quantizer for BER minimized relay selection.Simulation results show thatselection relaying with quantized CSI provides a good trade-off between the signalingoverhead and the system performance,and is very attractive for practicalimplementation.In a multiuser relaying system,wireless network coding is a more efficienttransmission scheme than conventional relaying schemes.However,it still suffersfrom spectral efficiency loss as compared with the point-to-point direct transmission.By exploiting the broadcast nature of the wireless medium and the linear property ofpractical error control codes,we propose a spectral efficient cooperative transmissionscheme for multiuser relaying systems based on joint network coding and channel decoding.Simulations results show that the proposed scheme outperforms thepoint-to-point direct transmission in terms of BER performance in various cases.