Current wireless systems are capacity constrained networks. Emerging wireless local area networks (WLANs) and home audio/visual networks demand near 1 Gb/s or more data rates to support multiple high-speed high-definition television (HDTV) A/V streams. Another challenge faced by these networks as well as outdoor wireless wide area network (WWAN) systems for fixed/nomadic access is non-line-of-sight (NLOS) propagation and multipath propagation due to scattering by obstacles (illustrated in Figure 1). These effects produce random fluctuations in signal strength known as fading. The received signal variations result in variation in signal to noise ratios (SNR) at receiver, which is very sensitive at both transmitter and receiver. In this section, all these issues are discussed in brief and more specifically the need of MIMO system is presented.

The theoretical channel capacity $(C)$ or throughput limits are defined by the Shannon - Hartley theorem as illustrated,

$C = B* log_2(1+\gamma)$

where, $B :$ Channel Bandwidth and $\gamma$ : Received SNR. This equation shows that in a wireless channel with Single - Input, Single - Output (SISO) as illustrated in Figure 2, the capacity up to 1Gbits/sec (or spectral efficiency of 9bits/sec/Hz) can be achieved by improving received SNR ($\gamma$) and increasing channel bandwidth. SISO is a traditional one – one antenna system at transmitter and receiver respectively. The effect of improving receiver SNR and increasing bandwidth on capacity of channel is discussed in the following section.

i) Increase in Capacity by improving received SNR $(\gamma)$ : Received SNR can be improved by,

a) Increasing transmitted power in a terminal:

From Shanon's Theorem, spectral efficiency (i.e. Capacity per bandwidth) of a wireless channel with an SNR of 10 dB (10) is $C / B = log_2(1+\gamma) = log_2(1+10) = 3.46 $ bits/sec/Hz. If power is increased by a factor of 10 times, i.e. SNR of 20dB will increase the spectral efficiency to 6.65 bits/sec/Hz.. Similarly, a 100 times increase in power, i.e. SNR of 30 dB will increase the spectral efficiency to 9.96 bits/sec/Hz, approximately a tripling of spectral efficiency. The capacity is increasing as a log function of the SNR, which is a slow increase. Clearly significant increase in received SNR results in marginal gains in channel capacity. However transmitted power can not be increased according to one's choice, as it acts as an interference to other users. Due to large usage of transmitted power, the battery may also get exhausted quickly. Generally, it is limited to less than 1W in indoor environments due to biohazard considerations. These limits are about a factor of ten higher in outdoor tower - based Base Stations.

b) Design of linear receiver with low phase noise: It is difficult to design a low cost linear receiver for wireless channel due to requirement of high power RF amplifiers. The peak SNR achievable limit is 30 - 35 dB in a wireless receiver.

In previous chapters, it is discussed that in cellular systems, high capacity and hence spectral efficiency can be achieved by frequency reuse concept. However co-channel interference, channel fading in the presence of imperfect power control and peak power limitations at the transmitter are the main problems associated with a cellular system. These factors result in the peak achievable SINR lower than the received SNR limit of 30 - 35dB. The average SINR in a cellular reuse scheme lies in the range of 10 - 20 dB at the best. This implies that increasing the spectral efficiency in a SISO NLOS cellular network beyond a peak value of 4 - 6 bits/sec/Hz (average value of 2 - 4 bits/sec/Hz) is not possible. In pure line-of-sight (LOS) links, practical SISO systems have attained spectral efficiencies up to 9 bits/sec/Hz. However, such systems rely on fixed point - to point links with very high gain directional antennas and Fresnel clearance to eliminate fading completely. The advantage of high-gain antennas in reducing the transmit power constraint is not available in NLOS environments, where large angle spread due to scattering can make such antennas highly inefficient.

Note: From, this discussion it is clear that data rates of 1 Gb/s can not be achieved just by increasing the power level by 100 times or with the use of high gain antennas in NLoS environment.

As a result, the other way to achieve higher data rates of 1 Gb/s or spectral efficiency more than 9 bits/sec/Hz is to increase the signal bandwidth $(B)$.

ii) Increase in Capacity by increase in signal bandwidth ($B$):

Consider a system that realizes a nominal spectral efficiency of 4 bits/sec/Hz over 250-MHz bandwidth, so that the data rate is 1 Gb/s. 200MHz of bandwidth is scarce, is not impossible to obtain, particularly in frequency bands below 6 GHz, where NLOS networks are feasible. 250 MHz of bandwidth is easier to obtain in the 40 - GHz frequency range. However at frequencies higher than 6 GHz, shadowing increases due to obstructions in the propagation path, which makes the NLOS links unusable. Due to limitations in transmitted power, a signal of 250 MHz bandwidth can not be used for longer distances. Another way is to increase the signal bandwidth of a communication channel by increasing the symbol rate of a modulated carrier along with coding. However it increases its susceptibility to multipath fading further puts limit on capacity.

Finally it is to be noted that Gb/s wireless links in NLOS (cellular) networks using conventional techniques are not feasible due to peak and average SNR or SINR limits in practical receivers. Additionally, high bandwidth systems put a limit on larger distances. Therefore, design of transmitter and receiver for very high data rate, say 1Gb/s wireless links that offers good quality-of-service (QoS) and range capability in NLOS environments is an engineering challenge. These challenges are addressed by exploiting the multipath propagation characteristics of wireless channel which itself is constraint in the capacity of wireless communication.