Advanced Techniques for Multiuser Wireless Systems

From Access Wiki

Contents 1 Overview 2 Participants 3 Activities 3.1 Multi-carrier Systems 3.2 Multiple-Relay Networks and Cooperative Communications 3.2.1 Overview 3.2.2 Main Problems Studied 3.2.3 Publications 3.3 Cognitive Radio Techniques 3.4 Sensor Networks 3.4.1 Overview 3.4.2 Publications

[edit] Overview

[edit] Participants

Giuseppe Cocco Ana Maria Galindo, Ana Moragrega, Moshbah Shaat, David Pubill, Faouzi Bader, Lorenza Giupponi, Christian Ibars, Marco Di Renzo

[edit] Activities

[edit] Multi-carrier Systems

---

[edit] Multiple-Relay Networks and Cooperative Communications

Participants: Aitor del Coso, Christian Ibars

[edit] Overview

Interest is growing in alternative communication strategies beyond the traditional multi-user schemes of wireless multiple access or multiplexing, because of the fundamental limitation of the fact that these schemes are characterized in information theory by the multiple access channel and broadcast channel respectively, and in both cases the capacity per node tends to zero as the number of nodes increases. In cooperative transmission, nodes cooperate in order to increase their overall throughput by, for example, relaying information or creating virtual arrays. Recent results have shown the potential of these techniques for application in future wireless systems, such as wireless mesh networks, sensor networks, or beyond 3G wireless access.

This field is justified by the fact that emerging wireless networks, such as community networks or sensor networks, involve multiple wireless hops. This trend is also supported by standardization groups, which are working in mesh networking support for 802.11 and 802.16 networks (.n and .j extensions, respectively). The main objectives of this line of research are: to provide efficient cooperation algorithms for multi-hop wireless transmission and cooperative wireless access; to explore the fundamental limits of wireless networks; and to integrate such algorithms into a wireless network architecture.

Multi-hop wireless networks may find application in all wireless systems. Current wireless networks such as WLAN or UMTS are basically access networks, where the wireless portion is limited to a single hop, and wired infrastructure is deployed to support it. For high capacity networks, where the cell size is small, the wired infrastructure can be very costly and difficult to deploy. A way to reduce wired infrastructure costs is to provide wireless links for some of the network interconnections beyond the access link. Additionally, for some applications other than wireless access, such as sensor networks, there may not be wired infrastructure at all.

[edit] Main Problems Studied

• The multiple-parallel relay channel:
  • Achievable rates with (partial)D&F, C&F, linear relaying.
  • Scaling laws and Asymptotic performance.
• The multiple-access channel with multiple-parallel relays:
  • Achievable rate regions with D&F, C&F and linear relaying.
  • Weighted sum-rate optimization
• The broadcast channel with multiple relays
  • Rate regions and duality

[edit] Publications

• Outage performance of a virtual MIMO channel in a Sensor Network (journal1)
• Achievable rates for the multiple relay channel: (partial) D&F (journal2, proc1), linear relaying (proc2 )
• MAC-BC duality for multiple relays (proc3)

(See authors' websites for a comprehensive list)

---

[edit] Cognitive Radio Techniques

---

[edit] Sensor Networks

Participants: Ana Moragrega, Aitor del Coso, Christian Ibars

[edit] Overview

The focus of this research line is the global optimization of network energy consumption. The key aspect is to come up with a realistic energy model for:

• operating states (transmit, receive, idle, sleep, ...)
• state transitions

A simple linear model for the power consumption of a transmitting node is

P_T = P_c + k₁P_rf

where P_c is a fixed term accounting for transmit circuitry consumption, k₁ > 1 accounts for the efficiency of the power amplifier, and P_rf is the transmitted RF power. A similar model can be defined for a receiving node (with only a constant factor).

Commercial sensors (e.g. Chipcon) follow a linear model in certain output power ranges, although others have limited power management capabilities.

[edit] Publications

• A cooperative transmission algorithm which clusters sensors to create virtual MIMO channels (journal1)