Ferkans — Interactive Telecom Tutor

Virtual MIMO from Single-Antenna Nodes

MIMO systems (Chapter 15) achieve spatial diversity by using multiple antennas at the transmitter or receiver. But many wireless devices — sensors, IoT nodes, low-cost handsets — have only a single antenna. Cooperative diversity creates a virtual MIMO array by having nearby single-antenna nodes share their antennas: each node overhears its partner's transmission and re-transmits the information, so the destination receives independent copies through spatially separated paths. The seminal work of Laneman, Tse, and Wornell (2004) showed that cooperative protocols can achieve the same diversity order as physical MIMO systems, fundamentally changing how we think about diversity in distributed networks.

Definition:
Cooperative Diversity and Virtual MIMO

In a cooperative diversity system, a source $S$ and one or more partner nodes (relays $R_1, \ldots, R_L$ ) cooperate to transmit information to a destination $D$ . The protocol operates in two phases:

Phase 1 (Broadcast): $S$ transmits its message. Both $D$ and the relays $R_1, \ldots, R_L$ receive noisy copies.

Phase 2 (Cooperation): Each relay that successfully decoded (or simply amplified) re-transmits to $D$ .

The destination combines all received copies using maximum-ratio combining (MRC) or a space-time code. The diversity order of the cooperative system is:

$d = L + 1$

where $L$ is the number of cooperating relays that forward, achieving the same diversity as an $(L+1)$ -antenna system.

Cooperative diversity sacrifices multiplexing gain (the half-duplex loss) for diversity gain. In the diversity-multiplexing trade-off (DMT) framework: $d(r) = (L+1)(1 - 2r)$ for repetition-coded cooperation, where the factor of 2 reflects the two-phase protocol.

Definition:
Dynamic Decode-and-Forward (DDF)

Dynamic decode-and-forward (DDF) is a cooperative protocol where the relay listens until it has accumulated enough information to decode, then switches to forwarding for the remaining block length. Let $f$ denote the fraction of the block used for listening. The relay decodes after fraction $f$ if:

$f \cdot \log_2(1 + \text{SNR}_{SR}) \geq R$

After decoding, the relay transmits for the remaining $(1-f)$ fraction using a space-time code with $S$ . The DDF protocol achieves the optimal DMT of the cooperative channel:

$d_{\text{DDF}}(r) = (L+1)(1 - r)$

which matches the DMT of a $(L+1) \times 1$ MISO channel with no half-duplex loss in the DMT sense.

Theorem: Diversity Order of Cooperative Protocols

Consider a cooperative system with one source, $L$ relays, and one destination, all with single antennas and i.i.d. Rayleigh fading links. The outage probability at high SNR behaves as:

Repetition-coded cooperation (fixed DF): $p_{\text{out}} \doteq \text{SNR}^{-(L+1)} \quad\text{(diversity order $L+1$, multiplexing gain $1/(L+1)$)}$

Space-time coded cooperation: $p_{\text{out}} \doteq \text{SNR}^{-(L+1)} \quad\text{(diversity order $L+1$, multiplexing gain $1/2$)}$

Selection relaying (relay forwards only if it decodes): $p_{\text{out}} \doteq \text{SNR}^{-(L+1)} \quad\text{(diversity order $L+1$, multiplexing gain $1/2$)}$

Dynamic decode-and-forward: $d_{\text{DDF}}(r) = (L+1)(1-r), \quad 0 \leq r \leq 1$

All protocols achieve full cooperative diversity order $L+1$ , but they differ in multiplexing gain. DDF achieves the optimal DMT of the $(L+1) \times 1$ MISO channel.

Each relay provides an independently faded copy of the source message, just like an additional antenna. The probability that all $L+1$ paths (direct + $L$ relay paths) are simultaneously in deep fade decays as $\text{SNR}^{-(L+1)}$ . The key insight is that even though the relays are geographically separated (not co-located antennas), the diversity benefit is the same as co-located MIMO — only the array gain differs.

Proof

Outage event characterisation

For selection DF cooperation, the system is in outage if either: (i) No relay decodes and the direct link is in outage, or (ii) Some relays decode but the combined link (direct + decoded relays) is in outage.

Let $\mathcal{D} \subseteq \{1, \ldots, L\}$ be the set of relays that decode. Relay $\ell$ decodes if $\log_2(1 + \text{SNR}|h_{SR_\ell}|^2) \geq 2R$ (accounting for the half-duplex factor).

Combined rate at destination

Given decoding set $\mathcal{D}$ , the destination combines the direct link and the relay links using MRC:

$\text{SINR}_{\text{MRC}} = \text{SNR}|h_{SD}|^2 + \sum_{\ell \in \mathcal{D}} \text{SNR}|h_{R_\ell D}|^2$

Outage occurs if $\log_2(1 + \text{SINR}_{\text{MRC}}) < 2R$ .

High-SNR diversity analysis

At high SNR, $\Pr[\text{relay } \ell \text{ fails to decode}] \doteq \text{SNR}^{-1}$ . By independence:

$\Pr[\text{all relays fail}] \doteq \text{SNR}^{-L}$

If at least one relay decodes, the MRC combiner has diversity $|\mathcal{D}| + 1$ , so:

$\Pr[\text{outage} \mid |\mathcal{D}| \geq 1] \doteq \text{SNR}^{-2}$ (at minimum).

The dominant outage event is all relays failing and the direct link failing:

$p_{\text{out}} \doteq \text{SNR}^{-L} \cdot \text{SNR}^{-1} = \text{SNR}^{-(L+1)}$

yielding diversity order $d = L + 1$ . $\blacksquare$

,

Cooperative Diversity-Multiplexing Tradeoff

Animated DMT curves comparing the direct link, orthogonal AF (reduced multiplexing gain), and DDF (optimal MISO DMT).

The DDF protocol achieves the optimal

2 \times 1

MISO DMT

d(r) = 2(1-r)

, while simpler orthogonal AF protocols suffer a multiplexing loss with

r_{\max} = 1/2

.

Cooperative Diversity Performance

Compare the outage probability of direct transmission, repetition- coded cooperation, selection relaying, and space-time coded cooperation as a function of SNR. Adjust the target rate to see how the multiplexing loss affects the absolute outage performance. All cooperative protocols achieve diversity order 2 (with one relay), visible as the steeper slope of the outage curve at high SNR compared to direct transmission (diversity 1).

Parameters

Target rate

R

(bits/s/Hz)1

Example: Cooperative Diversity Gain with One Relay

A source communicates with a destination via a direct Rayleigh fading link. A single relay is available for cooperation. The target rate is $R = 1$ bit/s/Hz and all links have the same average SNR.

(a) Compute the outage probability of direct transmission at SNR = 10 dB and SNR = 20 dB. (b) Compute the outage probability of selection DF cooperation at the same SNR values (assume half-duplex loss). (c) At what SNR does cooperation achieve $p_{\text{out}} = 10^{-3}$ ? (d) What is the cooperation gain (in dB) at $p_{\text{out}} = 10^{-3}$ ?

Solution

Direct transmission outage

(a) For Rayleigh fading: $p_{\text{out}}^{\text{direct}} = 1 - \exp(-(2^R - 1)/\text{SNR})$ .

At 10 dB ( $\text{SNR} = 10$ ): $p_{\text{out}} = 1 - e^{-1/10} = 1 - 0.905 = 0.095$ .

At 20 dB ( $\text{SNR} = 100$ ): $p_{\text{out}} = 1 - e^{-1/100} \approx 0.01$ .

Cooperative outage

(b) For selection DF with half-duplex (effective rate $2R$ ): $p_{\text{out}}^{\text{coop}} \doteq c \cdot \text{SNR}^{-2}$ .

At 20 dB: $p_{\text{out}} \approx c/100^2 = c \times 10^{-4}$ where $c \approx (2^{2R}-1)^2/2 = 9/2 = 4.5$ .

$p_{\text{out}}(20\text{ dB}) \approx 4.5 \times 10^{-4}$ .

At 10 dB: $p_{\text{out}} \approx 4.5/100 = 0.045$ .

SNR for target outage

(c) Direct: $p_{\text{out}} = 10^{-3}$ requires $\text{SNR} \approx 1/10^{-3} = 1000$ (30 dB).

Cooperative: $10^{-3} = 4.5 \times \text{SNR}^{-2}$ $\Rightarrow \text{SNR}^{2} = 4500$ $\Rightarrow \text{SNR} = 67.1$ (18.3 dB).

Cooperation gain

(d) Cooperation gain: $30 - 18.3 = 11.7$ dB.

The relay provides an 11.7 dB reduction in the SNR required to achieve $10^{-3}$ outage probability — this is the tangible benefit of diversity order 2 vs. 1. $\blacksquare$

Quick Check

What is the main advantage of cooperative diversity over physical MIMO diversity?

Cooperative diversity achieves higher diversity order than MIMO

It creates a virtual antenna array from geographically distributed single-antenna nodes, enabling diversity without multiple antennas per device

It eliminates the half-duplex loss completely

It requires no channel state information at any node

Correction:

It creates a virtual antenna array from geographically distributed single-antenna nodes, enabling diversity without multiple antennas per device

Cooperative diversity allows devices with a single antenna to achieve spatial diversity by sharing their antennas with neighbours. This is especially valuable for IoT, sensor networks, and low-cost devices that cannot accommodate multiple antennas.

Cooperative Diversity

A technique where multiple single-antenna nodes cooperate by relaying each other's messages, creating a virtual MIMO array. With $L$ cooperating relays, the system achieves diversity order $L + 1$ , matching the performance of an $(L+1)$ -antenna MISO channel in terms of outage probability slope at high SNR.

Virtual MIMO

A distributed antenna array formed by geographically separated single-antenna nodes that cooperate to transmit or receive a common message. Unlike co-located MIMO, virtual MIMO requires inter-node communication (relay links) and incurs a half-duplex penalty, but achieves the same diversity order as physical MIMO.

Related: Cooperative Diversity

Common Mistake: Ignoring the Half-Duplex Rate Penalty in Cooperative Diversity

Mistake:

"Cooperative diversity with $L$ relays achieves the same rate as an $(L+1)$ -antenna MISO system, with additional diversity gain."

Correction:

Cooperative diversity achieves the same diversity order as $(L+1)$ -antenna MISO, but not the same rate. Each relay must operate in half-duplex mode: it listens for a fraction $\alpha$ of the time and transmits for $(1-\alpha)$ . This introduces a multiplexing loss: the DMT of the orthogonal AF protocol is

$d(r) = (L + 1)\left(1 - \frac{r}{1/(L+1)}\right) = (L+1)(1 - (L+1)r),$

which has maximum multiplexing gain $r^* = 1/(L+1)$ instead of 1. Only the dynamic decode-and-forward (DDF) protocol achieves the optimal MISO DMT $d(r) = (L+1)(1 - r)$ , at the cost of variable-length listening phases that complicate implementation.

In practice: 5G NR IAB nodes operate in time-division duplex with a configurable DL/UL split, but the half-duplex loss is always present.

Historical Note: The Birth of Cooperative Communications

2000--2012

The idea that distributed single-antenna nodes could cooperate to create a virtual MIMO array emerged independently from several groups around 2000--2003. Sendonaris, Erkip, and Aazhang (2003) proposed "user cooperation diversity" for CDMA uplinks, showing that two mobiles could share their antennas. Laneman and Wornell (2003) introduced coded cooperation. The seminal 2004 paper by Laneman, Tse, and Wornell unified these ideas into a single framework, proving that selection relaying and space-time coded cooperation achieve full diversity order with simple protocols.

The diversity-multiplexing tradeoff analysis by Azarian, El Gamal, and Schniter (2005) revealed the fundamental half-duplex penalty and introduced the dynamic decode-and-forward (DDF) protocol as the information-theoretically optimal strategy.

Cooperative diversity became one of the most active research areas in wireless communications during 2003--2012, producing thousands of papers and influencing the design of LTE-Advanced relay nodes and eventually 5G NR IAB (Integrated Access and Backhaul).

Key Takeaway

Cooperative diversity = spatial diversity without multiple antennas. With $L$ cooperating relays, the system achieves diversity order $L + 1$ , matching an $(L+1)$ -antenna MISO in outage slope. The cost is a half-duplex multiplexing loss: simple protocols reduce the maximum multiplexing gain to $1/(L+1)$ , while the DDF protocol recovers the full MISO DMT $d(r) = (L+1)(1-r)$ at the cost of implementation complexity. Cooperative diversity is most valuable for devices that cannot support multiple antennas — IoT sensors, low-cost UEs, and relay-assisted coverage extension.

Cooperative Diversity

Virtual MIMO from Single-Antenna Nodes

Definition: Cooperative Diversity and Virtual MIMO

Definition: Dynamic Decode-and-Forward (DDF)

Theorem: Diversity Order of Cooperative Protocols

Outage event characterisation

Combined rate at destination

High-SNR diversity analysis

Cooperative Diversity-Multiplexing Tradeoff

Cooperative Diversity Performance

Parameters

Example: Cooperative Diversity Gain with One Relay

Direct transmission outage

Cooperative outage

SNR for target outage

Cooperation gain

Quick Check

Cooperative Diversity

Virtual MIMO

Common Mistake: Ignoring the Half-Duplex Rate Penalty in Cooperative Diversity

Historical Note: The Birth of Cooperative Communications

Key Takeaway

Definition:
Cooperative Diversity and Virtual MIMO

Definition:
Dynamic Decode-and-Forward (DDF)