Hierarchical constellation based data transmission schemes in multi-hop cellular systems

Abstract

Recently, multi-hop networks are being spotlighted for the next generation networks. The most attractive benefits of multi-hop communication systems are not only the enhancement of user capacity and service coverage but also the reduction of deployment cost. In this dissertation, we focus on the capacity enhancement through relay in multi-hop networks. By introducing relay systems, we can enhance signal to noise ratio and therefore we can acquire higher data rate between a transmitting node and a receiving node with given radio resources. However, without spatial reuse of radio resource and the spatial diversity gain induced by multi-path relay, this benefit is reduced or meaningless in multi-hop networks. According to the growing demand of broadcasting systems such as DVB systems, many researches on data transmission schemes using hierarchical constellation are progressing to increase link capacity. However, there are no data transmission scheme using the characteristics of hierarchical constellations in the case of unicast systems, especially for multi-hop networks. In this dissertation, we propose hierarchical constellation based data transmission schemes by utilizing the differing degrees of bit robustness against channel errors for multi-hop networks. Then, we examine the capacity gains of the proposed data transmission schemes through numerical analysis under multi-cell environments. At first, we propose an adaptive resource allocate scheme based on hierarchical constellations. A symbol is allocated to a user entirely in conventional scheme whereas we divide a symbol into several bits and allocate these bits to multiple users adaptively to the channel condition of each user in our scheme. Because a symbol consists of multiple bits with different robustness against channel errors, a bit-by-bit resource allocation can be easily applied for adaptive resource allocation. To maximize the total cell capacity, we formulate a maximization problem...