Uplink Sum Capacity

¡¡

The ergodic capacity of a point-to-point MIMO channel has been extensively studied in the past decade. As all the transmit signals experience the same large-scale fading, the ergodic capacity can be fully described as a function of the average received signal-to-noise ratio (SNR) ¦Ì, which does not vary with the positions of transmit/receive antennas as long as ¦Ì is a constant. In a single-user DAS, in contrast, paths between the user and BS antennas are subject to different levels of large-scale fading. In this case, in addition to the average received SNR ¦Ì, the ergodic capacity is further determined by the normalized large-scale fading vector ¦Â [Dai'11]. By spreading out the BS antennas, ¦Â has unequal entries, indicating that the channel dimensions are reduced compared to the case with colocated BS antennas. It was recently shown in [Dai'11] that for a given average received SNR ¦Ì, a lower ergodic capacity is achieved with distributed BS antennas due to a smaller receive diversity gain if full channel state information (CSI) is available only at the receivers. If the transmitters are able to track the instantaneous channel status, on the other hand, such a reduction of channel dimensions leads to enhanced channel fluctuations, which can be further exploited to provide better water-filling gains and a higher capacity over the co-located case in the low SNR region.

For the multi-user single-cell case, with multiple co-located BS antennas, the ergodic sum capacity of a vector multiple-access channel is a function of the average received SNRs of K users {¦Ì_k}. If BS antennas are distributed over the cell, on the other hand, the uplink ergodic sum capacity is further determined by the normalized large-scale fading vectors {¦Â_k} [Dai'11]. Both {¦Ì_k} and {¦Â_k} depend on the positions of users and BS antennas, which may lead to prohibitively high computational complexity when the number of BS antennas or users is large. To decouple the comparison of the uplink ergodic sum capacity with co-located and distributed BS antennas, uplink power control was considered in [Dai'11] where the transmission power of each user is properly adjusted to overcome the large-scale fading such that the average received SNR ¦Ì_k is a constant. By doing so, all the users achieve similar rate performance regardless of their positions, which is desirable in fairness-constrained or delay-limited scenarios. Similar to the single-user case, it was shown in [Dai'11] that for a given average received SNR, the sum capacity of a single-cell DAS is lower than that with co-located BS antennas if full CSI is available only at the receivers. With full CSI at both transmitters and receivers, the sum capacity is significantly improved and becomes superior to the co-located case thanks to better multi-user diversity gains.

For the multi-user multi-cell case, the performance is severely limited by intense inter-cell interference due to aggressive frequency reuse among cells. In [Dai'14], the comparative framework was further extended to a cellular system where Lc BS antennas are either co-located at the cell center or uniformly distributed within each cell. With a large number of users, the inter-cell interference density is shown to be inversely proportional to Lc if the co-located antenna (CA) layout is adopted. With the distributed antenna (DA) layout, it scales in the order of Lc^{-¦Á/2}, where ¦Á is the path-loss factor, and is much lower than that in the CA case when Lc is large. Thanks to the reduction of inter-cell interference level, a much higher uplink sum capacity is achieved by the DA layout even if CSI is absent at the transmitter side.

 

Lin Dai, "A Comparative Study on Uplink Sum Capacity with Co-located and Distributed Antennas," IEEE J. Sel. Areas in Commun., vol. 29, no. 6, pp. 1200-1213, June 2011.

Lin Dai, "An Uplink Capacity Analysis of the Distributed Antenna System (DAS): From Cellular DAS to DAS with Virtual Cells," IEEE Trans. Wireless Commun., vol. 13, no. 5, pp. 2717-2731, May 2014.