We discuss in this paper the effects of antenna efficiency in the presence of a user and on the performance of a multiple-input multiple-output system. The investigation is carried out in different usage scenarios where the antenna efficiency changes significantly from one scenario to another. The propagation channel is defined to include the 3D antenna electromagnetic properties with the 3GPP spatial channel model. We then investigate at a link level using a Long Term Evolution (LTE) link simulator where the antenna and the user are integrated into the link. The analyses are based on channel correlation, channel power transfer factor, and the link throughput. We show the importance of accurately accounting for the antenna efficiency for realistic MIMO link performance. We also show that accounting for the antenna efficiency is a weighted factor defined by the antenna pattern behavior.

The comparison of the outage capacity for multiple-input multiple-output (MIMO) channel models based on different underlying approaches is made. Three different channel models are considered: the 3GPP empirical spatial channel model (SCM), a multi-element transmit and receive antenna (METRA) analytical spatial channel model (A-SCM) and the correlation-based long-term evolution (LTE) channel model. The authors evaluate the models' predicted channel capacity for different antenna element separation, array orientation and angle spread, with and without mutual coupling. The authors compare these results with measurement campaigns from the literature. The authors also derive an effective distance term that combines the antenna element separation, array orientation and angle spread parameters. The authors use this value to describe the effect on the signal correlation of the antenna output, and thereby explain the outage capacity dependence on the variables. Among the considered channels, the SCM showed the best agreement with the measurement literature, followed by the A-SCM and then the LTE model. The SCM was also the most computationally involved, followed by the A-SCM and then the LTE model. Our analysis showed that the mutual coupling had a small impact on the performance of all channel models, especially for antenna element separations greater than half a wavelength.

We propose a hybrid non-orthogonal amplify- decode and forward (NADF) transmission protocol for two half- duplex relay nodes. This strategy allows simple amplify and forward (AF) nodes and more complex decode and forward (DF) nodes to participate jointly in the transmission of information. Such a strategy is useful in networks with heterogeneous devices. Analysis of the diversity-multiplexing trade-off (DMT) shows our protocol to be better than the dynamic decode and forward (DDF) protocol for rates between 1/2 and 2/3. Our technique demonstrates that the DDF protocol is not an upper bound and that protocols exist which perform closer to the DMT limit. The NADF protocol has equal performance to the DDF protocol for rates between 2/3 and one. It has a DMT performance below the DDF protocol and identical to or better than the slotted amplify and forward (SAF) protocol and non-orthogonal amplify and forward (NAF) schemes for rates from zero to 1/2.

We evaluate the multiple-input multiple-output (MIMO) channel capacity dependence on angular spread, antenna element separation, and angle of arrival. The empirically derived spatial channel model (SCM) is selected for the simulations and an uncoupled ideal dipole uniform linear array is considered. We show that for a small angular spread (8deg), the angle of arrival (AoA) will increase the required signal-to-noise ratio (SNR) for a high ergodic capacity by more than 4 dB as it moves away from the array broadside. For the case of a large angular spread (68deg), the AoA has a minimal impact on the ergodic capacity. Insight into the capacity behavior is explained in terms of the minimum channel eigenvalue and the path correlation. We also found that as the channel outage capacity threshold decreases, the impact of the system parameters is reduced.

The paper considers the single relay orthogonal cooperative diversity decode and forward scenario as a block fading channel model. A closed form expression for the pairwise error probability is derived which leads to design criteria for full diversity and coding gain. Furthermore, the power distribution between the source and relay units is introduced as a design parameter. Fixed power distribution allocation has the potential of improving performance with no increase in the system complexity. Several techniques for improving the coding gain are given based on the power distribution during the cooperation phase of the transmission. As an example, the (1,7/5) convolutional code was considered. Simulation results for this example show that a 1 dB performance gain can be achieved by modifying the power distribution of the source and relay. For the code considered, the equal power allocation method most commonly used gave the worst performance. The best performance was achieved when the source was silent during the cooperation phase. This observation simplifies protocol design.

A major challenge in the area of space time (ST) codes is to find codes suitable for efficient decoding, thus overcoming the problem of many existing ST code designs which require maximum-likelihood (ML) decoding. A solution could be to apply single-input single-output (SISO) channel codes and theory over temporal channel fading to the multi-input single-output (MISO) code construction and classical suboptimum decoding methods. For these purposes, an ST code construction which allows the use of efficient decoding algorithms is described. We propose a concatenated code, where the inner code is the diagonal ST Hadamard (D-STH) code with Paley constructions and the outer code is an algebraic block code, such as a Reed-Solomon (RS) code. As a decoding method, we investigate a bounded minimum distance (BMD) with erasure decoding algorithm. A simple method to achieve the optimum threshold for deciding on an erased symbol is derived. Using this, the proposed concatenated scheme is able to exploit almost all of the spatial diversity of the system.

The critical design criterion for space-time codes in asymptotically good channels is the minimum rank between codeword pairs. Rank codes are a two-dimensional matrix code construction where by the rank is the metric of merit. We look at the application of rank codes to space-time code design. In particular, we provide construction methods of full-rank codes over different complex signal constellations, for arbitrary numbers of antennas, and codeword periods. We also derive a Singleton-type bound on the rate of a code for the rank metric, and we show that rank codes satisfy this bound with equality.