Effects of Sectorization on Cellular Radio Systems Capacity with Different Traffic Loads
Wireless Personal Communications: An International Journal
Optimum scheduling for smart antenna systems in Rayleigh fading channel
MILCOM'03 Proceedings of the 2003 IEEE conference on Military communications - Volume I
| Hi-index | 0.07 |
This paper investigates a hybrid space-time-division multiple access (S-TDMA) for broadband indoor wireless systems using sectored antennas. It is shown that portables which are located in different sectors of an indoor microcell may be able to reuse the same frequency and the same time slot. However, this requires careful scheduling of packet transmissions in order to avoid transmitting packets that would jam each other during the same time slot. It is proposed that the scheduling be performed in the base station, i.e., a central control architecture. The optimum scheduling algorithm, the one that maximizes the number of packets transmitted per frame, may be in the NP-complete class of problems, so it cannot be solved in real time. Therefore, a suboptimum algorithm, called the first fit algorithm (FFA), is proposed for frame scheduling. It was found that the capacity gain achieved by the FFA is dependent on the capture threshold, which is defined as the minimum signal-to-interference ratio required in order to achieve a given packet error rate goal. The capture threshold depends on the modulation and coding schemes. This paper investigates the performance of the FFA operating with multicarrier trellis-coded modulation. An alternative multicarrier modulation is analyzed, and the FFA performance is investigated when operating with the alternative and with the conventional multicarrier through computer simulations based on measured data which were obtained with a sectorization level of ten (using ten antenna sectors in the base station). The simulations have shown that the FFA can provide a large capacity gain when operating with multicarrier trellis-coded modulation and using differential detection. For example, while previous schemes can transmit only one packet at a time, the proposed scheme can transmit, on average, more than four packets per time slot in an open indoor location, or close to three packets per time slot in a closed indoor location with internal walls of concrete blocks