||The wireless communications market is one of the most important economic markets in terms of investment. The constant demand for more services of wide applications and requirements has driven the market in one of its fastest evolutions. Communicating with anyone, anywhere, anytime as expected in future wireless systems has proven much stronger needs than what is being provided today in current systems. The need for extreme high data rates and quality of services in heterogeneous networks has made current transmission and detection techniques by far not suitable for the next wireless generations. Viable solutions to this problem include the use of antennas arrays or Multiple Input Multiple Output (MIMO) systems, as well as, powerful signal processing techniques. MIMO systems have been shown to provide unprecedented gains such as diversity or multiplexing rates. The latter is the number of its degrees of freedom and is closely related to its capacity. Most of the previously designed MIMO systems have focused on the maximization of either types of gain and have been categorized as either diversity type systems or high (data) rate ones. Orthogonal designs (OD) and Vertical-Bell Layered Space-Time schemes (V-BLAST) are such examples. By maximizing exclusively one type of gain, these schemes, unfortunately, sacrifice the other type such as the system capacity with OD and diversity with V-BLAST. Designing schemes that can maximize both types of gains is a challenge for future wireless systems, particularly when multiple arrays of antennas are expected to be deployed. The overall objective of this thesis is to design multiple antenna array schemes that achieve both high capacity and high diversity gains. To do so, we consider the use of group detection at the receiver. In multiple antenna arrays environments, signals from one array of antennas arrive at the receiver as one group. Group detection will then preserve their internal structure and its use is justified. We will consider two basic receivers' structures for group detection, a Group Zero Forcing (GZF) and a Group Decision Feedback Detector (GDFD). We will first investigate these receivers in the existence of improper multi-access interference. We will demonstrate that improvement of the detection rules in this case will lead to significant diversity gains. We then consider the application of the GDFD receiver in a multi-user MIMO environment. Particularly, we will show that ordering the received signals from the different arrays is a very important issue to maximize the overall system performance. Optimal and sub-optimal ordering algorithms will then be proposed. Recently, a new performance measure, known as the diversity-multiplexing tradeoff func-tion, has been proposed to measure the tradeoff between the diversity gain and the multiplexing one. Such function has been shown to be a key performance comparison and evaluation tool. This thesis will assess the performance of the GZF and GDFD receivers in terms of diversity-multiplexing tradeoff in a multi-user MIMO system over a richly scattered Rayleigh fading channel. The optimal tradeoff function will be derived and three rate allocation algorithms will be proposed accordingly; namely, equal rate, group size proportional rate and optimal rate allocation. Results will demonstrate how powerful group detection is when applied for multi-ple antenna arrays. In particular, it is shown that group detection achieves significantly high diversity gains and capacities than regular decorrelators and with much less complexity than the optimal receiver. Designing schemes that achieve the optimal performance of group detection will be the focus of the last part of this thesis. We will also propose transmission schemes that will inter-encode the different arrays of antennas with a very special structure allowing various attractive aspects. These include structure flexibility regarding the number of antennas, low complexity receivers such as with Space Time Block Codes, full diversity performance, and high data rata transmission. This inter-array coding will be denoted as Block Orthogonal Linear Dispersion (BOLD) codes and they will be regarded as a parent class of the Space-Time Block codes. Performance analysis will demonstrate unprecedented achieved diversity levels at very high data rates.