Deliverable 3

8Relaying studies for broadcast networks

Cooperative communications in wireless networks have recently been introduced as a new method to increase the throughput, the reliability and the robustness of wireless systems. Relay techniques in cooperative networks where relay nodes assist the transmission between a source and a destination have been widely explored in the literature, namely, the decode-and-forward [138], the amplify-and-forward [139] and the compress-and-forward [140] techniques.

Most commonly, relaying with time sharing between the source and the relay has been widely explored in the literature. However, time sharing is not an efficient solution for broadcast scenarios, i.e., it is not convenient that the base station transmits half the time and remains silent the other half. We propose two new approaches suitable for broadcast scenarios. The first approach is based on frequency sharing, where the available bandwidth is shared between the base station and the relay to achieve orthogonality at the receiver side. The second approach is a distributed coding/modulation scheme which allows at efficiently exploit the available resources (time, frequency).

The remainder of this part will be as follows. In Section 8.1, we provide coverage analysis based on achievable rates. We consider relays operating in both half-duplex and full-duplex modes. In Section 8.2, we present a distributed coding scheme suitable for broadcast scenarios with OFDM based transmission under the half-duplex constraint (frequency sharing). In Section 8.3, we present a new relaying strategy for full-duplex mode called decode-rotate-and-forward based on constellation rotation at the relay. The latter coding scheme has been filed as a patent and the application is under process.

8.1Coverage Extension through Cooperative Relaying

Radio signals traveling from the transmitter antenna to the receiver antenna exhibit attenuation due to the natural diffusion of the signal, absorption and diffraction phenomena. This attenuation called path-loss is generally assumed proportional to at least the square of the distance between the transmitter and the receiver. In wireless infrastructures, receivers can exhibit severe attenuation which results in low throughput, poor coverage and reliability. Enhanced capacity and extended coverage are among the desired requirements for wireless communication systems. Relaying offers a promising solution to improve reliability, achieve higher throughput or provide coverage extension. In this work, we investigate the effect of fixed relay deployment on coverage in a wireless infrastructure.

8.1.1Coverage Analysis for the Decode-and-Forward Strategy

For coverage study, as for outage analysis, we fix a target rate and seek to maximize the geographic region outside which an outage occurs. We compute the coverage area for the cooperative relay network.

_{Figure : Representation of the wireless relay network in two-dimension model}

We consider the wireless relay network in two dimensions as depicted in Figure . Without loss of generality, downlink transmission is considered in this work. A base station BS communicates data to a mobile station MS cooperating through a relay R. The relay is fixed and its location is given by coordinates, the coordinates of the base station are set to and the mobile station location is given by coordinates This two-dimension model allows to assess the performance of relay transmission under realistic deployment scenarios. The base station transmits with power P_BS and the relay with power P_R. We further assume that the relay is at best as powerful as the base station, i.e. P_R ≤ P_BS.

We denote by l_BS,MS the link between the base station and the mobile station, l_BS,R the link between the base station and the relay, and l_R,MS the link between the relay and the mobile station and we denote by γ_BS,MS, γ_BS,R and γ_R,MS the signal-to-noise ratio in dB of the base station-to-mobile station channel, the base station-to-relay channel, and the relay-to-mobile station channel, respectively. Likewise, we denote by d_BS,MS, d_BS,R= and d_R,MSthe distance from BS to MS, from BS to R, and from R to MS, respectively. The channels are modeled as additive white Gaussian noise (AWGN) channels with path loss attenuation. The mean power of the signals decreases with distance d as d^-α, where α denotes the path-loss exponent and it is often assumed to be 2≤α≤6. For propagation in free space, the path-loss exponent α is assumed to be 2. For lossy environments, α can be in the range of 4. The path-loss exponent is assumed to be constant over the two dimensional plane. Therefore, γ_BS,R is given by

where N denotes the noise power. Similar expressions are obtained for and.
Let R > 0 denote a desired transmission rate. We denote by G the coverage region. For direct transmission where a base station transmits without the help of the relay, the coverage region is given by

G

where C denotes the capacity of the direct link.

In the two dimensional plane, this is equivalent to find coordinates which verify (2). By solving (2), we have that the mobile station is out of coverage if

8.1.2Coverage Extension with Half-Duplex Mode

For the cooperative scheme, we recall that the achievable rate R' for the decode-and-forward strategy with relay operating in half-duplex mode [141] is given by

As described in Section 8.1.1, the relay position is fixed at distance from the base station. is chosen such that it verifies


If verifies (5), the relay is still able to decode the information transmitted by the base station.
For a fixed distance between the source and the relay, we define the coverage region as

G

This is equivalent to find the pair which verifies

In Figure , we plot the coverage area for both direct transmission and cooperative transmission. Here, we consider R=1/2, P_BS/N=P_R/N=2 dB and a path-loss exponent α =3.52.

As shown in Figure , the coverage is extended especially at the side where the relay is located. However, a loss in the coverage is observed with respect to direct transmission in some regions. This is due to the half-duplex constraint, i.e. the available resource is shared between the base station and the relay.

_{Figure : Coverage area of the base station for both direct and cooperative half-duplex relay transmission. R=1/2, PBS/N=PR/N=2 dB and α=3.52.}

8.1.3Coverage Extension with Full-Duplex Relay

The coverage study for the full-duplex relay scenario remains valid. The achievable rate of decode-and-forward relaying scheme is given by

The relay position is chosen such that it verifies

and the coverage area for a given distance is obtained by finding the coordinates which verify

In Figure , we plot the coverage area of the base station with and without the full-duplex relay cooperation. We assume a rate R=1/2, P_BS/N=P_R/N=2 dB and a path-loss exponent α=3.52.

_{Figure : Coverage area of the base station for both direct and cooperative full-duplex relay transmission. R=1/2, PBS/N=PR/N=2 dB and α=3.52.}

Unlike the half-duplex case, the coverage is always extended compared to direct transmission. While full-duplex relaying outperforms half-duplex relaying in terms of coverage, full-duplex scenario requires additional complexity at the receiver since signals from both base station and relay are received simultaneously and also at the relay to tackle potential interference between the relay transmitting and receiving antennas [142]. The choice between half-duplex and full-duplex relaying should be made in a compromise between coverage extension and complexity. In the next sections, we propose distributed coding schemes with relay operating in both half-duplex and full-duplex modes.