No abstract

Abstract-An efficient scheme for the multiple-access multipleinput multiple-output (MIMO) channel is proposed, which operates well also in the single user regime, as well as in a direct-sequence spread-spectrum (DS-CDMA) setting. The design features scalability and is of limited complexity. The system employs optimized low-density parity-check (LDPC) codes and an efficient iterative (belief propagation-BP) detection which combines linear minimum mean-square error (LMMSE) detection and iterative interference cancellation (IC). This combination is found to be necessary for efficient operation in high system loads 1. An asymptotic density evolution (DE) is used to optimize the degree polynomials of the underlining LDPC code, and thresholds as close as 0.77 dB to the channel capacity are evident for a system load of 2. Replacing the LMMSE with the complex individually optimal multiuser detector (IO-MUD) further improves the performance up to 0.14 dB from the capacity. Comparing the thresholds of a good single-user LDPC code to the multiuser optimized LDPC code, both over the above multiuser channel, reveals a surprising 8-dB difference, emphasizing thus the necessity of optimizing the code. The asymptotic analysis of the proposed scheme is verified by simulations of finite systems, which reveal meaningful differences between the performances of MIMO systems with single and multiple users and demonstrate performance similar to previously reported techniques, but with higher system loads, and significantly lower receiver complexity.Index Terms-Code-division multiple access (CDMA), iterative decoding, low-density parity-check (LDPC) code, multiple-input multiple-output (MIMO) channel, multiuser.

No abstract

No abstract

Abstract-In this paper we investigate the achievable rate of a system that includes a nomadic transmitter with several antennas, which is received by multiple agents, each with a single antenna, suffering independent channel coefficients and additive Gaussian noises. Since the transmitter is nomadic, the agents do not have any decoding ability. These agents process their channel observations and forward it to the final destination through lossless links with a fixed given capacity. Assuming Gaussian signalling, we get lower and upper bounds on the achievable rates, and demonstrate the achievability of the full multiplexing gain. We also extend the model to address multi-user systems. The asymptotic setting with numbers of agents and transmitter's antennas taken to infinity is examined, and the incompetence of the simple compression when compared to a Wyner-Ziv scheme is demonstrated. For finite setting, an upper-bound is derived, which turns out to be quite tight when compared to the WynerZiv achievable rate, even for a rather small 4 × 4 system. I. INTRODUCTIONIn this paper we deal with a network setting in which a nomadic transmitter has several antennas and is communicating to a remote destination, where no direct link exists between the transmitter and the final destination. The final destination receives all of its inputs from several separated agents, which are connected to it through lossless links with a given capacity. The channel between the transmitting antennas and the agents is the standard ergodic Rayleigh fast fading channel with independent fading. The channel state is known to the agents and the final destination, but not to the transmitter. Since the transmitter is nomadic, the agents do not possess the codebook in use, and thus do not have any decoding ability [1]. This setting is closely related to the setting of the Multiple input multiple output (MIMO) channel, which is thoroughly treated in the literature, see [2] and others. We focus here on the multiplexing gain [3], which is a typical characterizing feature for MIMO systems. The results here have also implications on more complicated channels that include MIMO, such as the MIMO broadcast channel [4], the MIMO relay channel [5], and ad-hoc network [6]. All these works deal with situations where multiple antennas are transmitting and are received in a distributed fashion, either by relays, destinations or any combination of the above. In addition, results regarding adhoc networks [7], relay channels [8], and joint processing [9] are closely related, providing another aspect of the achievable rates in wireless networks, where relays form, in a distributed manner the required spacial dimensions. This paper is also linked to source coding problems, since we limit the agents to process only source related algorithms, such as compression. Relevant works are e.g. [10], who deals with the multiple Wyner-Ziv problem, the Gaussian CEO solution by [11] and