Performance Comparisons of Fixed and Adaptive Beamforming Techniques for 4G Smart Antennas Adnan Anwar Awan†⁕1, Irfanullah†1, Shahid Khattak†2, Aqdas Naveed Malik⁕3 †
COMSATS Institute of Information Technology, Abbottabad, Pakistan 1
[email protected], 1
[email protected], 2
[email protected] ⁕ ISRA University I-10 Markaz Islamabad, Pakistan. 3
[email protected]
Abstract—Much attention has been paid to the design and implementation of adaptive antenna arrays (also called smart antennas), intended for use in current and future wireless communication systems. The motivation is that employment of such array antennas can significantly increase the range and capacity of wireless systems by orienting the main beam towards the desired direction while simultaneously placing nulls toward the interferers (thereby increasing the signal to interference SIR ratio). In this paper, various adaptive beamforming algorithms have been compared based on nulls location, gain and half-power beamwidth (HPBW). Performance comparisons of matrix inversion beamforming algorithm (used for fixed beamforming) and LMS, RLS (used for adaptive beamforming) are investigated in terms of their effects on main beam scanning, resolution and convergence. In particular, the effects of nulls location on the gain and half power beamwidth (HPBW) of main beam have been investigated for the fixed vs adaptive beamforming algorithms. Keywords— Smart Antenna, Beamforming, LMS, RLS, HPBW, Gain
I.
INTRODUCTION
In modern wireless communication systems, beamforming is the widely used technique that improves antenna gain significantly and reduces the interferences. The analog devices used in traditional beamforming requires an expensive hardware resulting in highly expensive solution. Digital beamforming done by software significantly increases the flexibility and decreases the extra cost expended on the hardware [1]. The development of wireless technology networks undergoes different generations (1G, 2G, 3G, 4G and 5G). 4G is the most advanced generation in mobile communications networks and is still looking for feasible and optimized solutions for suitable spectrum access, better power control , inter and intra base station interference cancellation. Hence there are many opportunities available for advance research utilizing beamforming in these networks to increase the network capacity. Exponential increase in users and their demands for various wireless services needs to have a wider coverage area and higher transmission quality. The solution to this problem is a „smart antenna‟. A great deal of research is already been done on adaptive antennas algorithms [2-8]. In mobile communications the demand for wireless services is growing exponentially day by day resulting in insufficient number of frequencies to support the users. So cell sectorization [9] is used to provide more frequencies per coverage area. Despite of the benefits of cell sectoring it is not sufficient enough to provide solution to capacity problem now a days. Therefore designers are thinking
of dynamic sectorization of the cells. So smart antennas are ideal in this regard. Smart antennas are basically the extension of cell sectorization in which the coverage of the sector is composed of multiple beams [9].As smart antennas can focus their beam pattern towards the signal of interest (SOI) and mitigate the signal not of interest (SNOI). Therefore they could provide a large coverage area. Besides this as smart antennas have tendency of higher rejection of SNOIs can therefore have low bit error rate (BER), resulting in improved substantial capacity. Theproposed work is organized in the following manner: In section II, the concept of smart antenna is presented, section III discusses digital beamforming, section IV is about beamforming algorithms, then section V presents simulation results and finally section VI concludes the paper. II.
THE CONCEPT OF SMART ANTENNA
Smart antenna system refers to an antenna array which is capable of adjusting its beam pattern according to the changes in electromagnetic surroundings. The adjustment is done using sophisticated signal processing techniques and algorithms. The purpose of adaptation is to enhance the Signal Of Interest (SOI) and to minimize the signal in which we are not interested (Not Signal Of Interest (SNOI)) known as the interferers in the literature [13]. The sophisticated digital processing involves digital beamforming. Smart antennas are of two types, one switched beamforming systems, in which several beam patterns are available and the system is switched to the suitable beam on the basis of requirements of the system and the other one is adaptive beamforming system, which steers the main beam to SOI while nulling the SNOIs at the same time. Smart antenna has several advantages over the traditional antenna which is not intelligent enough to sense the changes in environment and react accordingly. In mobile communications system it is now possible for a base station by using smart antenna to direct the narrow beam towards the SOI user and nulls toward the SNOIs user resulting in the higher signal to interference ratio. Hence the transmit power requirement will be low and provide higher tendency for frequency reuse within a cell resulting in increased capacity. III.
DIGITAL BEAMFORMING
Traditional antenna arrays are known as phased arrays, in which the main beam is steered to SOI by phase shifters. Modern beam steering or beamforming is done by the smart antennas which are referred as digital beamformed arrays (DBF). As mentioned above, the smart antennas are capable of producing suitable beam patterns for enhancing the SOI
according to the change in electromagnetic environment. These beam patterns are controlled by algorithms which are based on certain criteria. The criteria is steering main beam towards SOI and nulling towards SNOI. The implementation of these algorithms is done by using digital signal processing. The requirement of digital signal processing is that the array outputs are digitize by using A/D convertors. Digital beamforming has many applications among which radar system [10-14], sonar system [15] and communications systems [16] are few of them. There are two types of digital beamforming based on the angle of arrival (AOA) from the transmitters. When angles of arrival do not change with time, the optimum array weights required to drive the antenna array elements to produce the desired beamforming pattern need not to be adjusted. This type of beamforming is known as fixed beamforming. Whereas if the wireless channel is changing with time, the optimum array weights needs to be changed accordingly. This needs to compute the optimum array weights in real time. Such type of beamforming is called adaptive beamforming. IV.
BEAMFORMING ALGORITHMS
In order to implement smart antenna beamforming algorithms, an array of antennas is formed so that the array weights are applied on individual antenna elements in such a way that they enhance the main beam towards SOI and to mitigate nulls towards SNOIs. In this work, the performance of three beamforming algorithms namely Matrix Inversion Algorithm (MI), Least Mean Square (LMS) and Recursive Least Squares (RLS) have been investigated on eight elements linear antenna array . The details of these algorithms are as follows. A. Matrix Inversion (MI) Algorithm MI algorithm is used for fixed beamforming in smart antenna systems. In order to compute the complex weights wn for the proposed array to produce the required beam patterns, linear algebra method in [13] has been used. The complex weights are computed as: -1
wMI n =A B
(1)
-1 -3 -5 -7 7 5 [e 2 jkd sinθ e 2 jkd sinθ e 2 jkd sinθ e 2 jkd sinθ e2jkd sinθ e2jkd sinθ
]
-3
-5
-7
AF=w-4 e 2 jkd sinθ +w-3 e 2 jkd sinθ +w-2 e 2 jkd sinθ + w-1 e 2 jkd sinθ + 5
0