2202 IEICE TRANS. COMMUN., VOL.E88–B, NO.5 MAY 2005 LETTER Theoretical Study of Site Selection Diversity Transmission in DS-CDMA Cellular Mobile Radio Mahbub ALAM † , Student Member, Eisuke KUDOH † a) , and Fumiyuki ADACHI † , Members SUMMARY Single cell reuse of the same frequency, which is possible in DS-CDMA cellular systems, yields the option of site diversity to increase link capacity. In this letter, a generalized case of site diversity transmission is considered where multiple base stations (BS’s) are involved in weighted transmissions with constant total transmit power to a target mobile station (MS). A general equation of conditional bit error rate (BER) is derived based on the model of weighted transmissions combined with antenna di- versity reception and rake combining. It turns out theoretically that the optimum set of weights to maximize forward link capacity makes site se- lection diversity transmission (SSDT) the best performer. This theoretical analysis is confirmed by performance evaluation based on the Monte-Carlo simulation. key words: DS-CDMA, site diversity, soft handoff, forward link capacity, rake combining 1. Introduction Wideband DS-CDMA is used for multiple access scheme in 3rd generation cellular systems due to its high spectrum efficiency. In cellular systems, the service area is divided into many cells, where each of the cells has a base station (BS) in the cell center to serve users in the respective cell [1]. However, the distance dependent path loss, shadowing and fast fading cause a large fluctuation of received signal level at a mobile station (MS). The probability of the received signal power failing to achieve the required signal quality or bit error rate (BER) is called outage probability. Near the cell’s boundaries, the outage probability is high. However, remembering that the signals transmitted from two or more BS’s are received at nearly equal levels near a cell boundary, the same data is sent simultaneously towards a target MS from two or more BS’s and is combined by the MS receiver to improve the forward link transmission performance. This technique is called site diversity and is incorporated into soft handoff (SHO) which is helpful in a handoff transition from one cell to another [2]. Weighted transmissions from multiple BS’s involved in site diversity operation are expected to render further im- provement in transmission performance. This is because, there may exist a set of optimum weights that will opti- mize site diversity operation. In this letter, therefore, we first present the theoretical analysis, in which we assume weighted transmissions from multiple BS’s to a target MS. Manuscript received June 30, 2004. Manuscript revised December 2, 2004. † The authors are with the Dept. of Electrical and Communi- cation Engineering, Graduate School of Engineering, Tohoku Uni- versity, Sendai-shi, 980-8579 Japan. a) E-mail: kudoh@mobile.ecei.tohoku.ac.jp DOI: 10.1093/ietcom/e88–b.5.2202 After deriving the set of optimum weights that maximizes the received signal-to-noise plus interference power ratio (SINR), we show that the optimal solution to maximize the forward link capacity with constant total transmit power is to transmit only from the best BS that has the maximum channel gain, in other words it is SSDT as proposed in [3]. Furthermore, performance evaluation based on the Monte Carlo simulation confirms that SSDT gives better perfor- mance than SHO in terms of link capacity which is defined as the maximum number of users that can communicate si- multaneously under the allowable outage probability. Some other studies of SSDT have been found in the open litera- ture [4],[5]. In [4], SSDT, SHO and hard handover (HHO) are compared using BER as a performance measure, while in [5], the complexities that arise from the implementation of SSDT in a practical system are addressed. Unlike the above mentioned works, this letter studies macro-diversity transmissions in general by theoretical analysis and in the end finds SSDT rendering the best performance and thus it contributes a theoretical basis to SSDT. The remainder of this letter is organized as follows. Sect. 2 describes the system model under consideration. The theoretical analysis is presented in Sect. 3. In Sect. 4, the performance evaluation is done by the Monte-Carlo simu- lation for various system parameters. Finally, this letter is concluded with Sect. 5. 2. System Model 2.1 Cellular Structure and Propagation Model It is assumed that the service area consists of 19 identical hexagonal cells as illustrated in Fig. 1. The BS is located at the center of each cell. We do not consider cell sectoriza- tion and an omni transmit antenna is assumed at each BS. In Fig. 1, i and j denote cell and MS respectively where, j=0th MS is the target MS. In mobile wireless communication, the propagation channel can be modeled as the product of distance depen- dent path loss, log-normally distributed shadowing loss and multipath fading channel gain. It is assumed that the fad- ing channel consists of L discrete propagation paths having time delays (τ l for the lth path) of integer multiple of DS- CDMA chip duration, each being subjected to independent Rayleigh fading and that rake combining can resolve all L paths to coherently combine them based on maximal ratio combining (MRC) method [6]. The instantaneous received Copyright c  2005 The Institute of Electronics, Information and Communication Engineers