Communication Link Model using Trellis Coded Modulation for 5G Communication in the Tropical Regions

Signal fading is a major problem in the tropical regions. Here the signals get highly attenuated for torrential rain, hence the signal power diminishes and the bit error rate (BER) increases. To enhance the signal to noise ratio (SNR) or to minimize the BER a communication link model using trellis coded modulation (TCM) for 5G communication has been proposed in this paper for the tropical regions. Various diversity techniques have been applied to minimize BER in the tropical regions. STBC (Space-Time Block Coding) MIMO technique is a spatial diversity technique which has been used in the proposed model in this paper. Bit error rate values are obtained for MIMO multipath fading distributions viz. Rayleigh fading distribution and Rician fading distribution. BER values are also attained in respect of different MIMO techniques (2x2, 3x3 and 4x4). The 4x4 STBC MIMO System has been generated using hardware co-simulation block and the results are implemented in Xilinx kintex-7 FPGA board. From the analysis of the result it is observed that Rayleigh fading distribution reflects better performance than that of Rician fading distribution. 4x4 MIMO Technique renders better result in comparison with that of lower MIMO techniques.


Rain attenuation in tropical regions
Total precipitation [13] [14] has been calculated on the basis of the amount of attenuation that arises out of rainfall per unit area. The amount of attenuation per unit volume is defined as the Specific Rain Attenuation measured at [dB / km]. Calculation of specific rainfall is detected especially in relation to the degree of rainfall with symptoms like storms and electric waves at that particular place. The characteristics of rain embrace precipitation and raindrop size whereas the characteristics of the electromagnetic waves comprise the frequency band, the polarization type and the propagation index. The acquisition of a particular rainfall is further propagated by the propagating electromagnetic waves. Its impact changes relying on temperature of water, ratio of water and frequency belts of operation. Thus, the electromagnetic wave is absorbed by the rain and gets dissipated as heat and the remaining portion of the electromagnetic wave which is not absorbed is scattered in all the directions [15] as displayed in Fig.2.This scattered electromagnetic wave emits invalid signals or interferes with the input of the received signal and as a result of this, extra attenuation originates. If the wave propagation is short and comes close to the size of the raindrop, the absorption of the wave becomes higher and the propagation becomes greater. As a result, the absorption and propagation of the electromagnetic wave depend on the operating frequency, the rainfall rate, the composition of the rainfall and the size of the raindrop. Therefore, the estimation of precipitation depends on the rainfall rate and frequency regression coefficients. Here, the specific rain attenuation is considered as γ and represented numerically in Equation 1 given below:- In equation (1) stated above R denotes rain rate and both k and ∝ denote power law parameters which rely on different factors like distribution of raindrop size, temperature, frequency of operation and emission of electromagnetic radiation. The specified attenuation of rainfall is based on the category of polarization of electromagnetic waves because of round shape nature of raindrops. The attenuation rate which arises from the vertical polarization waves is smaller than that which is caused by horizontal polarization [16] [17]. The values of parameters k and ∝ have been reflected in ITU Rec. P.838-3 [18]. The available rainfall area shown in Equation 1 above illustrates the distribution of rainfall ranging one kilometer while the length of the total rainfall area between the transmitter and the receiver is calculated by the frequency of the particular rainfall γ and by the actual length of the path Lac if it is observed that rainfall has been scattered uniformly. If the rainfall is not uniformly scattered along the length of the wireless path, the calculation of rain attenuation based on the length of the actual path produces undesired results. In view of what is stated above, the horizontal uniformity of precipitation has been chosen. This incident has been identified as the effective length of path of the connection between the transmitter and the receiver and this effective length of path must be less than the actual wireless path length. This above phenomenon has been applied to assess the actual path length between a transmitter and a receiver for the uniform spreading out of rainfall. On the basis of the above phenomenon, the length of the effective path has been accepted as an important factor to analyze the rain attenuation. To compute the effective path length Leff, the minimization of the distance has been considered to measure the actual length of the path in case of a uniform spreading out of rainfall. So, a new metric has been suggested and it is defined as a distance deduction factor "dr" used to compute the effective path length between a transmitter and a receiver. The effective path length calculated in terms of mathematics is established on the basis of the length of the sphere itself Lac, between two sides of the interaction and the assumptions about the structure of the rainfall equation for the structural distribution. Therefore, the effective path length has been mathematically demonstrated in Equation 2 [18] [19] [20] below: Leff = dr Lac (2) In the equation (2) above Leff denotes the length of the effective path, dr indicates the distance deduction factor or distance factor as recommended by ITU [21] and Lac represents the actual length of the link between a transmitter and a receiver. Different kinds of significant predicted models have been designed and suggested by different researchers in different times to assess the horizontal rainfall variability in the tropical regions [22] [23]. The most popular model used in the tropical region is ITU-R model. This model has been analyzed as an outcome of frequency with rain rate 0.01% of the time, the executant in the specified attenuation model and the real path length. Although the ITU-R model has been used to find out the amount of precipitation fluctuations, it is not the appropriate model for the tropical regions. Therefore, a number of models such as updated Silva Mello model [24], Abdulrahman's model [25], and Lin's model [26] have been matured and used to assess the path deduction factor.

Trellis coded modulation
The trellis codes for band-limited channels arise out of combining convolution coding with modulation. This state of combining is itself called as trellis coded modulation (TCM). This form of signaling comprises two basic features-the first feature is that the number of signal points lying in the constellation used is greater than what is used for the modulation format of interest with the identical data rate. Here the surplus points permit excessiveness for forward error control coding without giving up bandwidth. The second feature is that convolutional coding is utilized to bring in certain reliability between successive signal points such that only certain patterns or sequence of signal points are allowed.
In communication system, the channel volume curves [27] suggest that doubling the number of generated signals improve the reliability of digital communication system over an additive white Gaussian channel by maximum 6dB. In other words, if the kb bits are to be transmitted, then using the letters A′ of 2kb + 1 and utilizing the proper encoding scheme, it is likely to significantly increase the effectiveness of digital communication system against the extra noise. This is performed by providing a standard computer installer at the rate R = Kb / (Kb + 1) and binary input data and then mapping the result (Kb + 1) of the code words to the alphabet. This framework was originally proposed by Ungerboeck in 1982 [28]. Again the TCM functions contain the Trellis code and the constellation mapper as shown in Figure 3. TCM includes the coder functions of rate R = Kb / Kb + 1 and an M-Ary display showing M = 2 kb input maps to a large constellation of M = 2 kb+1 point locations.

Fig.3. A simple Trellis Code Modulation
When a sequence of emitted signals {Sb} get polluted by additive white Gaussian noise (AWGN), the decoder starts beginning to create erroneous decisions, perhaps being engaged in a trellis path that is separated by more than one transmission from the real path. Such errors are called incident-events. When Viterbi decoding is used the probability of such errors are estimated at very high signal-to-noise-ratio (SNR) and the probability of such errors P ′ is mathematically expressed in the following manner: Where Q′denotes the Gaussian error integral. NEfree denotes the average number of errors with distance dE free. The free distance dEfree of the trellis refers to the minimum Euclidean distance between pairs of sequences {S b} and {Sb′} that the encoder generates and it is mathematically displayed below- On the other hand, the effective path length is computed using equation (2) for trellis coded modulation technique.
As the trellis coded modulation is an effective and efficient bandwidth technique obtained using convolution codes, it preserves bandwidth by measuring the number of points displayed. In this way the rate of bits is enhanced but the rate of symbols remains the same. Coding conversion permits the modification of symbols to create sequential codes. In trellis coded modulation the increasing of the constellation size decreases the Euclidean distance between constellation points but the sequence of coding provides a coding gain that overmatches the disadvantage of power which moves towards the level of higher constellation. Performance is measured by the benefit of encoding over a fixed signal. The Euclidean sorting criterion is not considered as the Hamming range.
By applying trellis coded modulation with STBC MIMO system, energy efficiency throughput and spectral efficiency are obtained. Besides, by using trellis coded modulation with MIMO system normalized utility power and normalized packet delay are attained. So steps have been put in STBC MIMO system.

STBC MIMO System
Here a code word has been taken into account and a 4x4 MIMO system has been created with that code word [29]. This 4x4 MIMO STBC (Space time block coding) caused by 2x2 Alamouti STBC to receive diversity is developed using a building code adopted by Walsh-Hadamard to detect error and code. Here four sets of symbols and four transmitting and receiving antennas have been adopted. The same concept of transmitting signals from transmitters is applicable to the construction of the 4x4 Alamouti STBC. Thereafter the symbols are twisted sharply either by mixing at random or by changing the phase and then the symbols are again transferred. This method is repeated for the other two spaces, and later this process is repeated for all information. Now it is felt necessary to mention some advantages of MIMO. MIMO is extensively applicable to increase the capacity of channels by exercising multiple antenna configurations. In the MIMO system the data rate is improved by applying spatial multiplexing but the authenticity of the system is developed by space time coding. Space time block codes have been adopted in MIMO system for enhancing the diversity gain. Let xt1, xt2, xt3, and xt4 are the symbols which are to be transmitted. In the first slot, the said four symbols are transmitted in fundamental order. In the second slot, coded symbols such as xt2 *, -xt1 *, xt4 * and -xt3 * are transferred consequently. Similarly, xt3 *, xt4 *, -xt1 * & -xt2 * and xt4, -xt3, -xt2 & xt1 are transmitted in the third and fourth time Slots. The negative signs and conjugates are used here to suggest a phase shift for supporting orthogonality which is the main property of QOSTBC (Quasi orthogonal space time block code). In the same way, the information is transmitted from four transmitting antennas in four time intervals. Now it is necessary to create 4x4 STBC transmission matrix using Walsh-Hadamard repetitive composition rule on 2x2 matrix of Alamouti transmission. The steps below describe the style of transference of coded symbols to obtain diversity.

Proposed model
The communication link model has been implemented using hardware co-simulation block. The description of the proposed model and the hardware design has been depicted in sections 6.1 and 6.2 respectively.

Description of the proposed model
In order to minimize BER and to increase SNR the following model shown in Fig.5 has been proposed for 5G communication for the tropical regions. This model has been designed using Xilinx System generator. The Bernoulli binary generator generates samples as a vector output. The samples pass through the unbuffer block which converts a frame to scalar samples output at a higher rate. The input samples then pass through the 'Gateway In' block. This block converts the inputs of different types of Simulink integers to Xilinx fixed-point data. Then the data goes through 'Gateway Out'. This block converts Xilinx fixed point inputs into outputs of Simulink integers. The output data now passes through the buffer block. The buffer converts scalar samples to a frame output at a lower rate. The input samples now pass through the TCM encoder designed for tropical regions. An automatic gain controller (AGC) is added in the TCM encoder. This encoder convolutionally encodes binary data and modulates using the phase shift keying method. This encoder is designed in such a way that when signal passes through this encoder the signal power remains high. Then the input data passes through OSTBC encoder which is devised for tropical regions. This encoder encodes the input message using an orthogonal space-time block code (OSTBC). The OSTBC comprises rate 1 for 2 transmit antennas, rate 1/2 or 3/4 for 3 and 4 transmit antennas. This encoder is designed in such a way that when input message passes through this encoder then its SNR remains high. The input signal now passes through MIMO channel. The MIMO channel filters input signal through a MIMO multipath fading channel. In the MIMO channel there are included two types of fading channels viz. Rayleigh fading and Rician fading. From MIMO channel the signal passes through the rain attenuation channel. When a signal passes through this rain attenuation channel, the signal gets attenuated. Now this input signal makes its way through the OSTBC combiner. This combiner combines the received signal and channel estimate inputs in accordance with the structure of the orthogonal space-time block code (OSTBC). This combiner is drawn in such a manner that when the input signal goes through this combiner, the signal power gets enhanced. Therefore, the input signal moves along TCM decoder. This decoder uses the Viterbi algorithm to decode trellis-coded modulation data, modulated using the phase shift keying modulation method. Thereafter the input signal travels along the error rate calculation which computes the error rate of the received data by comparing it to a delayed version of the transmitted data. The block output is a three-element vector consisting of error rate, followed by the number of errors detected and the total number of symbols compared. This vector can be sent to either the workspace or an output port.
The values of BER relies on a number of parameters like no. of samples per frame, no of transmitting antennas in the transmitter and no of receiving antennas in the receiver. The BER values also depend on types of fading distribution.

Table 1: Change of BER due to various parameters
Hence by applying equations (6)- (13) and (15)-(18) the 4x4 OSTBC MIMO System used in the proposed communication link model has been designed using system generator and it is displayed in Fig. 6 given below. This OSTBC MIMO system is much effective in the tropical regions because it reduces signal attenuation and enhances BER of the signal. This OSTBC MIMO System has been reflected in the communication link model for 5G Communication using system generator demonstrated in Figure 5.  (12).The HDL code has been employed to implement the models. The RTL Schematic, top level RTL Schematic, Schematic view, designed model using FPGA editor, device utilization summary have been tabulated below in table 3. In table3 utilization of resources as displayed appears much less. Results for low cost and low power System on Chip (SoC) in consequence with a 4x4 OSTBC MIMO has been designed as highlighted in tables 3 and 4..Maximum delay of the clock is 1.231000 ns and net skew is 0.135000 ns.

Result analysis and discussion
The result of the overall work of the proposed model has been analyzed in this paper. It reveals that the analysis of result provides two parts-simulation part and synthesis part. After synthesis the reconfigurable block has been implemented using Xilinx Kintex7 FPGA board and its device Utilization is shown in Table 3.In sec 7.1, the result of the simulation part has been reflected and in sec 7.2, synthesis part has been demonstrated. In the synthesis portion timely behavioral analysis is important for two purposes. Firstly, it is inevitable to verify the connection between two connectors whether the cycle used meets all the time requirements so that all the events are produced within the required time expected by the regional environment. Secondly, for integrating digital circuits, the delays are determined within which the output events must be produced. While the proposed model is simulated, the values of BER are exhibited on the display block of the model for Rayleigh fading distribution and Rician fading distribution. Hence the BER values have been compared with two different types of fading distributions and this phenomenon has been demonstrated in Fig. 13. Different types of MIMO techniques (2x2, 3x3 and 4x4) have been applied to this proposed model. The BER values for different MIMO techniques have been displayed on the display block of this model and the comparison of BER for different MIMO techniques have been plotted in Fig.14. In the figures 15 to 17, the synthesis part of the result illustrates the hardware data sheet of encoder, digital clock manager, and multiple clocks. In the figures 18 to 21 the same result depicts the hardware data sheet of timing diagram for encoder, digital clock manager and multiple clocks respectively.

Conclusion
In this paper this proposed model has been designed especially for the tropical regions. This model will satisfy the long cherished desire of the people of tropical regions in respect of wireless communication. This proposed model has been depicted in order to maintain minimum BER in the tropical regions. From figure 13 it is observed that Rayleigh distribution has provided the minimum bit error rate (BER) than that of Rician distribution. From figure 14 it would transpire that 4x4 MIMO technique has produced the better result than that of 2x2 and 3x3 MIMO techniques. In other words, by using 4x4 MIMO techniques the signal to noise ratio (SNR) increases in the tropical regions. For this enhancement of SNR the distortion of signal becomes much less if any major message is transmitted from a transmitter to a receiver. In the figures 15 to 21 the synthesis results of the communication link model have been reflected. The designed hardware model of 4x4 OSTBC MIMO System produces low cost and low power as shown in table 3, table 4 and it will be a very effective model for tropical regions for reducing signal attenuation. From the overall work performed in this paper it is concluded that the using of this proposed model will bring down the cost, power and increase the reliability of the system than that of using other models. As a result, this proposed model will be easily accessible to any user of the tropical regions.