On the effects of pilot-aided channel estimation on the capacity of QoS-aware heterogeneous OFDMA networks

Abstract

The deployment of heterogeneous networks (Het-Nets) is a way to increase the capacity of future mobile networks. However, the use of pilot symbols for channel estimation, which is needed to perform demodulation at the receiver, reduces spectral and power efficiency, and the presence of channel estimation error degrades the capacity of each link. In this paper we study the effects of pilot-aided channel estimation on the capacity of a two-tier HetNet with QoS-aware centralized admission control (AC) and radio resource allocation (RRA) algorithms that control the effects of cross-interference and optimize a service provider revenue. With this aim, we design and implement an AC algorithm that takes into account link-rate degradation due to channel estimation error and pilot-symbols overhead. The results for both uplink and downlink show that the impact of channel estimation error is not negligible and is especially relevant for downlink cellular users. However, the results also show that this degradation can be partially mitigated by allocating more power over pilot symbols.