End-to-End Joint Antenna Selection Strategy and Distributed Compress and Forward Strategy for Relay Channels

Rahul Vaze, Robert W. Heath Jr.

TL;DR

This paper tackles the fundamental tradeoff between reliability and data rate in multi-hop wireless relay networks. The authors introduce two solutions: JEEMAS, an amplify-and-forward antenna selection scheme with minimal feedback that outperforms state-of-the-art distributed space-time codes across multiple hops, and a Distributed Compress-and-Forward (CF) strategy that provably achieves the theoretical upper bound of the diversity-multiplexing tradeoff (DMT) for two-hop networks with arbitrary numbers of relays.

What Problem Does This Paper Address?

In cooperative wireless networks, relay nodes help transmit information from a source to a destination. Network performance is fundamentally characterized by the Diversity-Multiplexing Tradeoff (DMT): the balance between reliability (diversity gain, $d$) and transmission rate scaling (multiplexing gain, $r$).

Prior approaches suffered from two major roadblocks:

  1. Multi-hop channels ($N > 2$ hops): Distributed Space-Time Block Codes (DSTBCs) fail to achieve optimal DMT at intermediate multiplexing gains and introduce high encoding/decoding complexity, latency, and noise accumulation.
  2. Two-hop multi-relay channels: While Compress-and-Forward (CF) was known to achieve the optimal DMT for a single relay, existing schemes for multiple relays required complicated inter-relay partial decoding, creating analytically intractable rate expressions with unknown outage exponents.

Methods

The paper develops two complementary communication strategies:

  • Joint End-to-End Multiple Antenna Selection (JEEMAS):
  • Evaluated on an $N$-hop network where relays operate in full-duplex with amplify-and-forward (AF).
  • The destination uses channel state information (CSI) to identify an optimal end-to-end "path" of $m$ antennas per stage that maximizes end-to-end mutual information.
  • Employs a hybrid mode: it chooses $m = 1$ antenna per stage at $r = 0$ (maximizing diversity) and $m = \min_n M_n$ antennas per stage for $r > 0$ (maximizing multiplexing).
  • Relies on successive decoding at the destination and low-rate feedback ($\log_2(\kappa + 1)$ bits) to notify nodes which antennas to activate.
  • Distributed Compress-and-Forward (CF):
  • Designed for a two-hop relay channel with direct source-destination links and $K$ relay nodes.
  • Relays compress received signals using Wyner-Ziv source coding without decoding each other's transmissions.
  • Compressed observations are sent over the multiple-access channel to the destination via correlated source transmission techniques, followed by backward decoding at the destination.

Datasets

  • Not discussed in the paper. The work is purely theoretical and validated analytically using standard independent and identically distributed (i.i.d.) flat Rayleigh block-fading channel models.

Key Insights and Ideas

  • Inter-relay decoding is unnecessary for optimal DMT: Prior multi-relay CF frameworks assumed relays must partially decode one another to exploit correlation. The authors prove that independent Wyner-Ziv compression is sufficient to meet the optimal DMT upper bound, as the required quantization noise level does not scale with transmit power ($P$).
  • Path selection over complex distributed coding: Instead of spreading signals across all nodes with complex space-time coding—which accumulates noise at each hop—selecting a subset of independent paths preserves SNR and yields higher diversity gain at intermediate rates.
  • Two operational modes suffice: Rather than computing a distinct network antenna partition for every target multiplexing rate $r$, a simple two-state strategy ($m=1$ for $r=0$ and maximal $m$ for $r>0$) achieves both extreme corner points and provides near-optimal intermediate performance.

Key Results

  • Optimal 2-Hop DMT: The proposed distributed CF scheme achieves the information-theoretic cut-set upper bound for any number of relay nodes: $$d(r) = \min\left\{(M_0 - r)(M_1 + M_2 - r),\, (M_0 + M_1 - r)(M_2 - r)\right\}$$ for $r = 0, 1, \dots, \min\{M_0, M_2\}$.

  • Superior Intermediate Diversity: For multi-hop configurations where antenna counts across stages are not divisible by one another (e.g., $M_0 = 3, M_1 = 5, M_2 = 3$), JEEMAS yields strictly better diversity gain than the best-known DSTBC partitioning schemes.

  • Low Overhead: JEEMAS achieves its gains with minimal feedback overhead: only $\log_2(\kappa + 1)$ bits are needed from the destination, where $\kappa$ is the number of independent network paths.

Limitations and Future Work

  • Generalization Beyond Two Hops for CF: While the distributed CF strategy achieves the optimal DMT bound for two-hop topologies, extending its outage probability exponent analysis to general $N$-hop multi-relay networks remains an open, non-trivial challenge.
  • Full-Duplex and CSI Assumptions: JEEMAS assumes full-duplex relays with perfect frame-level synchronization and relies on the destination having complete CSI of all active candidate links to compute optimal paths.
  • Suboptimality Gap: Although JEEMAS hits the extreme DMT corner points and beats DSTBCs, it still does not fully close the gap to the theoretical cut-set upper bound at intermediate multiplexing gains for $N > 2$.

Summary updated

Abstract

Multihop relay channels use multiple relay stages, each with multiple relay nodes, to facilitate communication between a source and destination. Previously, distributed space-time codes were proposed to maximize the achievable diversity-multiplexing tradeoff; however, they fail to achieve all the points of the optimal diversity-multiplexing tradeoff. In the presence of a low-rate feedback link from the destination to each relay stage and the source, this paper proposes an end-to-end antenna selection (EEAS) strategy as an alternative to distributed space-time codes. The EEAS strategy uses a subset of antennas of each relay stage for transmission of the source signal to the destination with amplifying and forwarding at each relay stage. The subsets are chosen such that they maximize the end-to-end mutual information at the destination. The EEAS strategy achieves the corner points of the optimal diversity-multiplexing tradeoff (corresponding to maximum diversity gain and maximum multiplexing gain) and achieves better diversity gain at intermediate values of multiplexing gain, versus the best-known distributed space-time coding strategies. A distributed compress and forward (CF) strategy is also proposed to achieve all points of the optimal diversity-multiplexing tradeoff for a two-hop relay channel with multiple relay nodes.
BibTeX
@article{vaze2009end,
  author = {Rahul Vaze and Robert W. Heath Jr.},
  title = {{End-to-End Joint Antenna Selection Strategy and Distributed Compress and Forward Strategy for Relay Channels}},
  journal = {EURASIP J. Wirel. Commun. Netw.},
  year = {2009},
  url = {https://openreview.net/forum?id=iauXW40Ts7},
}

Loading related content...

Figure