HIGHLY EFFICIENT MULTIHOP RELAY TOPOLOGIES

The big challenge for broadband wireless system design comes up with the right balance between capacity and coverage that offers good quality and reliability at a reasonable cost. It is important to look at system spectral efficiency more broadly to include the notion of coverage area. Results presented in previous sections have demonstrated the high potential benefits for relay deployment with radio resource sharing, in terms of interference, MIMO combination, and multiuser transmission. To implement the radio resource reuse and achieve highly efficient relay deployments, appropriate frequency reuse and multiuser access strategies are required. Relay systems must be based on a topology that fully exploits effective resource assignment based on the spatial separation of nodes. In this section, we propose directional distributed relay for highly efficient multiuser transmission with reduced demands on radio resource.

Figure 1 depicts the directional distributed relaying architecture. This is based on a paired radio resource transmission scheme, and it is possible to achieve one radio resource to one user (or one group of users) in average, even with multihop relay. The radio resource can be defined as either frequency (e.g., subcarriers in an OFDMA symbol) or time (e.g., OFDMA time slots). Transmissions in the BS coverage are the same as the IEEE 802.16e standard. For relay links, paired transmissions are applied, where the BS forms two directional beams, or uses two sector antennas to communicate with RS1 and RS2 simultaneously. A paired radio resources are required: f₁ and f₂. The first radio resource (f₁) is applied to the RS–BS1 link and also to the RS2–MS links (in the RS2 coverage); while the second resource (f₂) is applied to the BS–RS2 link and also to the RS1–MS links (in the RS1 coverage). Radio resources are shared between the RSs and MSs. Each end-user employs a single pair of radio resources, on average.

 

Figure 1: Directional distributed relaying with paired radio resource.

Using the sharing scheme outlined above the interference can be controlled at the BS and RS nodes. In this relay configuration there are only two sets of interference, as also illustrated in Figure 1. The interference between the BS and MS groups (I₁ and I₂) can be detected and controlled by the BS. First, the BS could employ an adaptive array to exploit the spatial separation of the groups. Second, since the received power by each MS in each MS group is known to the BS, the BS can apply interference avoidance  between the two groups based on measured signal to interference plus noise ratio (SINR) and power control, where the transmit power of the two RSs are controlled for balancing the SINR according to the service requirement. Furthermore, in this scenario the expected level of interference is small because the BS connects to the MSs through a relay, which means the relay SNR-gain will be much higher than the SNR_access level. Interference between RSs (I₃ and I₄) can be reduced by array processing (including the use of sector antennas) at the RSs. Interference measurement for the efficient resource assignment can be achieved during the neighborhood discovery procedure. To achieve high levels of SINR (e.g., 10–25 dB), array processing, including the use of sector antennas at the RS, is desirable.

This proposed topology is fully compatible with the existing 802.16e standard and no modifications are required at MSs. Alternative deployments topologies are also possible based on the same concept, such as a single RS to cover a coverage hole. In such cases, the radio resource sharing is performed between the RS and its BS. It could be complicated for statistical studies as the performance is fully dependent on the deployment scenario. However, it is much more feasible in a realistic application environment by employing real channel measurements and ray tracers.