No abstract

No abstract

The comparison of beam selection and antenna selection techniques in conjuction with the multiple-inputmultiple-output (MIMO) architecture is provided for measured channels. The channels, captured on a 3D MIMO measurement system based on the virtual antenna array technology, are centered at 5.8 GHz with a frequency span of 500 MHz. Both flat-and frequency-selective fading are considered. The selection techniques can be applied to either the transmit end or both ends of the communication links, and both cases are considered. Shannon capacity is employed as the comparison criterion throughout the paper. The presence of the line-of-sight (LOS), and the correlation of interference are shown to impact the performances of both kinds of selection techniques. The use of beams does not eliminate the need for stream control, which is the regulation of the number of data streams when MIMO links interfere. The results show that in the indoor environment, the beam selection is only slightly better than the antenna selection in both narrowband and wideband channels when stream control is used. However, without stream control, beam selection offers a significant improvement over antenna selection.

No abstract

A bs t r a d 111 many applications of digital communication engineering, the hits of the data sequence transmitt,ed do not have eqnal significance. To seek a more reliable information transmission. some unequal error protection (UEP) coding schemes have been devisrd. In this paper, however, we do not suggest a new UEP coding scheme, instead, through intimate romhination of quantizer, interleaver, and the best convolutional code (CC) or trellis-coded modillation (TCM), the lJEP capabilit,y can be achieved. Moreover, to investigate t h e performance of the proposed UEP coding scheme, we deduce the bit error rates of the corresponding encoder input ports from the transfer function of the encoder. The transfer function is derived based on the state diagram, such that the l1EP capability of t h e encoder can be proved. Finally, from the sirniilation results we find that the mean square error between the original qiiant.ized signal and the received qnantized signal can be rednced when the presented method i s employed compared t,o the CC or TCM without the inter-1~ a.ve r 8B.2.1