Project topics and materials
Get Complete Project Material File(s) Now! »
Cooperative Diversity in Virtual MIMO Systems
If MIMO technology provides such significant performance gains, due to cost, complexity or other constraints the deployment of multiple antennas at user terminals in a network can be prohibitive. As an alternative, cooperative relaying has been suggested as enabler of spatial diversity in the face of the aforementioned limitations. The basic idea behind this concept stems from the broadcast nature of the wireless environment where the transmitted signal by a source can be overheard by as many receiving nodes as there might exist. Some of these nodes can play the role of relays that cooperate with the source in order to form a virtual MIMO leading to either transmit or receive diversity generation. Importantly, the cooperative diversity gain resulting from relaying can be leveraged into a multiplexing gain as we address a more balanced tradeoff between the two.
Coupling MIMO with Temporal and Frequency Diversities
Despite the resulting diversities a MIMO system has to offer, there also exists other types of diversity that can be coupled with traditional or virtual MIMO systems. For instance, the use of temporal and frequency diversities with MIMO have led to the invention of space-time and space-frequency codes (see [2] and references therein). Also, the joint design of ARQ re-transmission protocols with MIMO results in a substantial increase in the transmission reliability [3] by exploring the ARQ delay, i.e. the number of ARQ rounds, into a diver-sity gain. Cooperative MIMO relaying with ARQ could bring interesting enhancements to eventually both diversity and multiplexing gains. In some situations, relays may act as packet re-transmitters giving rise to a generalization of the classical ARQ mechanisms. This provides a substantial increase in the diversity gain compared to conventional ARQ espe-cially over a long-term quasi-static ARQ channel. In general, it is quite difficult to exploit all forms of diversity simultaneously due to conflicting system demands and constraints. Nevertheless, a combination of some of these forms may be sufficient to attain our design objectives in terms of latency, reliability and coverage.
Towards a Balanced Latency-Reliability Tradeoff
In [4], a unified performance evaluation framework where different relaying protocols were introduced for relay-aided wireless networks. These protocols can straightforwardly be extended to the MIMO multi-relay cooperation case. They can be classified into three half-duplex protocols named as protocol I, protocol II and protocol III yielding homologous system models as in the traditional MIMO, SIMO and MISO schemes, respectively. In protocol I, the first relaying slot is dedicated to the source node to broadcast its single while the relay and the destination are listening. During the second relaying slot, both the relay and source nodes transmit to the destination node. Protocols I and II are alike in the first time slot, but only the relay which engages in the second relaying-hop transmission to the destination node. Protocol III is essentially similar to protocol I except that the destination node does not receive from the source during the first time-slot. Note that protocol I realizes maximum degrees of broadcasting and receive collision because of the non-orthogonal multiple access it results in during the second relaying slot. Herein, we distinguish between two famous types of signal processing at the relay: DF where the relay attempts to decode the received signal replica then forwards it to the destination and amplify-and-forward (AF) where the relay only transmits a normalized-to-the-symbol-energy version of the received signal to the destination node. Irrespective of the signal processing used in both cooperation phases at each node, any half-duplex relaying protocol must fall within one of the three categories we have just described. Full-duplex relaying is beyond the scope in our study.
Low-Latency Cooperative MIMO Relaying with ARQ
In cooperative MIMO relaying, the reliability is gained at the expense of an increased latency due to the half-duplex constraint at the relay nodes. Different methods have been proposed to recover this loss. For instance, successive relaying using repetition coding has been introduced in [5] for frequency-flat fading relay channels. In [6], relay selection methods have also been proposed for cooperative communication with decode-and-forward relaying. A prominent alternative to reducing the throughput loss in half-duplex cooperative relaying is the combination of both ARQ and relaying. This approach would significantly reduce the half-duplex multiplexing loss by activating ARQ for rare erroneously decoded data packets, when they occur. Approaches targeting the joint design of ARQ and relaying in one common protocol have received more interest (see for instance [7]–[9]). Motivated by the above suggestion, we investigate spectrum-efficient cooperative transmission techniques where both ARQ and relaying are jointly designed.
• In chapter 2, we focus on single-carrier MIMO broadband cooperative transmissions with AF relaying. Most of the research work that has been carried out in this area has focused on frequency-flat fading channels. Relaying techniques for frequency-selective fading channels have recently been investigated by some authors (see for in-stance [10, 11, 12]). Among our contributions in this thesis manuscript, we propose spectrum-efficient relaying protocols where the multiplexing loss due to the half-duplex transmission constraint is reduced while providing interesting outage performance. We introduce two new relay ARQ protocols where the total time required for transmitting one data packet is significantly reduced compared with conventional relaying methods. We also evaluate both the outage error probability and average throughput performance of the proposed schemes, and show that they outperform classical cooperative MIMO relaying schemes.
Receiver Design of the MIMO-ARQ Relay Channel
It is crucial in MIMO wireless communication systems to design receiver schemes that re-cover the information sent by the transmitter with the lowest probability of error. The op-timal receiver implements either the maximum a-posteriori (MAP) or maximum-likelihood (ML) criteria. In which case, its computational complexity gets involved as our MIMO system gets large. Therefore, we mostly resort to sub-optimal receivers that guarantee close-to-optimal performance results but with lower complexity. Pertaining to our pro-posed spectrum-efficient MIMO relay ARQ with slot-mapping reversal (SMR) transmission strategy in chapter 2, our next contribution in the same chapter is described as follows.
• In chapter 2, we also investigate practical turbo receiver design that addresses the aforementioned considerations while alleviating the computational load that may arise when the number of ARQ rounds increases. Inspired by the concept of turbo packet combining, initially introduced by Ait-Idir et al. [13] and then extended in [14] for broadband MIMO ARQ systems in the presence of co-channel interference, we pro-pose to perform signal-level sub-packet combining at the destination node jointly over both time slots and multiple ARQ rounds. Then, we introduce a recursive over-the-ARQ-rounds implementation of this combining strategy, that considerably reduces its computational complexity. Compared with conventional ARQ-based cooperative relay-ing protocols, the proposed turbo receiver scheme is assessed in terms of the average throughput and its superiority is quite remarkable over the entire SNR region.
Cognitive MIMO Relaying: Practical Challenges and Solutions
As pointed out earlier, MIMO relaying can serve the secondary system in different ways. One approach is the design of cooperative beamforming or space-time block coding (STBC) schemes that sort out the dilemma of coexistence on the same spectrum [35]-[37]. Inevitably, this approach requires large feedback overhead and additional complexity to compute the beamforming and precoding matrices. Besides, its performance is proportional to the num-ber of radio-frequency (RF) chains that whenever increased it becomes costly to implement especially at the user equipment side. Another approach, simple and less expensive yet real-izes a good tradeoff among performance, cost and complexity, is TAS. In its simplest form, only the RF chain causing less interference on the primary system and enabling greater sec-ondary system performance is selected. TAS has been adopted in the LTE uplink (Rel. 8/9) and we adopt it herein as a promising technology candidate for beyond 5G massive-oriented MIMO systems [38]. If the MRC is applied at the receiver side, the technique is referred to as TAS/MRC. If selection combining is instead applied at the receiver side, the technique is referred to as TAS/SC. According to [39], TAS/MRC performs better than TAS/SC at the expanse of an increased complexity which does not pose a real burden if the MRC is implemented at the base station level.
Table of contents :
Acknowledgement
Abstract
R´esum´e
List of Publications
1 Motivation and Basic Notions
1.1 From MIMO to Virtual MIMO
1.1.1 MIMO Advantages
1.1.2 Cooperative Diversity in Virtual MIMO Systems
1.1.3 Coupling MIMO with Temporal and Frequency Diversities
1.2 Towards a Balanced Latency-Reliability Tradeoff
1.2.1 Relaying Protocols
1.2.2 Low-Latency Cooperative MIMO Relaying with ARQ
1.2.3 Receiver Design of the MIMO-ARQ Relay Channel
1.3 Cognitive Radio: Towards an Optimal Spectrum-Efficiency
1.3.1 Approaches for Improved Spectrum Sharing Performance
1.3.2 Cognitive Cooperative Relaying
1.3.3 Spectrum-Efficient Cognitive SIMO Relaying
1.4 Cognitive MIMO Relaying: Practical Challenges and Solutions
1.4.1 Cross-interference Mitigation in Cognitive MIMO Relaying with TAS/MRC
1.4.2 Point-to-Point Cognitive MIMO Systems with TAS/MRC
1.4.3 Cognitive MIMO Relaying with TAS/MRC
2 Single-Carrier MIMO-ISI Relay ARQ Transmissions
2.1 Introduction
2.2 Relay ARQ Sub-Packet Transmission Model
2.2.1 MIMO Relay ARQ System Model
2.2.2 Relay ARQ with SMR Strategy
2.2.3 Sub-Packet ARQ Transmission Model
2.3 Outage Probability and Average Throughput
2.3.1 Outage Probability
2.3.2 Average System Throughput
2.3.3 Simulation Results
2.3.3.1 Scenario 1: Relay close to Source (lSR = 0.3)
2.3.3.2 Scenario 2: Relay close to Destination (lSR = 0.7)
2.4 Turbo Packet Combining Receiver Design
2.4.1 Brief Description of the Concept
2.4.2 Soft Sub-Packet Combiner Derivation
2.4.3 Recursive Implementation
2.5 Simulation Results and Remarks
3 Cognitive SIMO Relaying: An Exact Outage Analysis
3.1 Introduction
3.2 Framework Description
3.2.1 Proposed System Model
3.2.2 Secondary System Transmit Power Model
3.3 Relaying Protocol and Received SNR Statistics
3.3.1 Scenario 1 – Complete I-CSI Acquisition
3.3.2 Scenario 2 – Partial I-CSI Acquisition
3.4 Exact and Asymptotic Outage Analysis
3.4.1 Outage Probability
3.4.2 Diversity-and-Multiplexing Tradeoff
3.4.3 Fixed Outage Constraint
3.4.4 Proportional Outage Constraint
3.5 Simulation Results and Remarks
3.5.1 Network Geometry
3.5.2 Simulation Results and Remarks
4 TAS Strategies for Cognitive MIMO Relaying
4.1 Introduction
4.2 Framework Description
4.2.1 System Model
4.2.2 Power Allocation for S-Tx and R
4.2.2.1 Fixed Interference Threshold (Qi = Qi)
4.2.2.2 Adaptive Interference Threshold
4.3 TAS/MRC Strategies for Cognitive MIMO DF Relaying
4.3.1 Relaying Protocol and MRC-Combined SINRs
4.3.2 SNR-driven TAS/MRC Strategy
4.3.3 SINR-driven TAS/MRC Strategy
4.4 Direct Transmission Outage Performance
4.4.1 Received SINR Statistics for the SNR-driven TAS/MRC
4.4.1.1 CDF Derivation of s→ss˙1
4.4.1.2 PDF Derivation of s→s s˙1
4.4.2 Direct Transmission Outage Probability for the SNR-driven TAS/MRC
4.4.3 Received SINR Statistics for the SINR-driven TAS/MRC
4.4.3.1 CDF Derivation of s→s s˙1
4.4.3.2 PDF Derivation of s→s s˙1
4.4.4 Direct Transmission Outage Probability for the SINR-driven TAS/MRC
4.5 End-to-End Transmission Outage Probability
4.5.1 Derivation of A1
4.5.2 Derivation of A3
4.5.3 Derivation of A2 for the SNR-driven TAS Strategy
4.5.4 Derivation of A2 for the SINR-driven TAS Strategy
4.6 Simulation Results and Implementation Prospects
4.6.1 PDF of the First-Hop Received SINR at S-Rx for both TAS/MRC strategies and Some Insights on Approximation (4.34)
4.6.2 First-Hop Outage Probability
4.6.2.1 Impact of Antenna Configuration
4.6.2.2 Impact of the Second-Order Statistic ps
4.6.3 End-to-End Outage Performance
4.6.3.1 Impact of Antenna Configuration
4.6.3.2 Impact of Relay Location
4.6.4 Implementation Prospects of both TAS/MRC Strategies
4.6.4.1 Antenna Index Feedback Load
4.6.4.2 CSI Acquisition and Antenna Selection
4.6.4.3 TAS in Future Wireless Communications
5 Conclusion
A Proof of Proposition 2
B Proof of Lemma 1
C Proof of Lemma 2
D Proof of theorem 1
E Derivation of Eq. (4.25)
F Derivation of 2 x and 2
G Proof of Theorem 2
Bibliography
