VLSI Implementation of Green Computing Control Unit on Zynq FPGA for Green Communication

The issue of the energy shortage is a ﬀ ecting the entire planet. This is occurring because of massive population and industry growth around the world. As a result


Introduction
There have been many issues with the scarcity of natural resources in the Earth because of the rapid population expansion and industrialization in the world [1]. Thus, people are worried about the future generation saving of those resources. This can be done using green technology of connectivity and energy-efficient machines [2]. The work represents a step in the direction of promoting green networking technology and energy-efficient devices. A control unit (CU), to minimize the power consumption, is installed on field programmable gate array (FPGA) in this work. A control unit is a part of a circuitry that regulates activity in the computer [3]. It gives instructions on how to react to instructions that the program sends to these devices in the logic unit, memory, input, and output devices [4]. Figure 1 shows the block scheme for the control unit. It selects and retrieves instructions from the main memory in the proper sequence and interprets them to allow other functional elements to perform the respective operations at the appropriate time [5][6][7]. Each input data is passed via the main memory to a processing device, comprising the four basic arithmetic functions (i.e., adding, subtracting, multiplying, and dividing) as well as certain logical operations such as data comparison and the selection of the required problem-solving method or a suitable alternative, on the basis of default decision criteria [8,9].
Of all these features and vast application in the field of computing, CU is regarded as one of the suitable components which can be used for green computing as well as green communication; also, these green computing and green communication makes the environment sustainable. By reducing its power consumption, the CU can be made suitable for green applications. Therefore, the power-efficient CU will be the great choice for communication technologies [10,11]. The power consumption of CU is optimized by its realization on FPGA devices. FPGAs are those devices which are made-up with semiconductor materials. It is called field programmable because it can be reconfigured/reprogrammed after its manufacturing [12][13][14]. FPGA devices are made-up with many components, and theses components are regarded as building components such as clock buffers (BUFGs), flipflops (FF), input/output (IO) ports, memory blocks (MB), and digital signal processors (DSPs) [15,16]. The building components of the FPGA device are represented in Figure 2. GC is also regarded as green information technology (green IT). In a broad way, GC term can also be coined as the method of designing, manufacturing, implementing, using, and disposing the mobile and computing peripherals and devices with least damage on environment resources. A brief idea of green computing is described in Figures 3 and 4. Figure 3 shows the things which are associated with green computing. These are such power management of devices, designing of energy-efficient devices, processors, and other computer devices, cloud, and virtualization. In cloud and virtualization, we try to communicate the data with cloud server and access the data from cloud. Figure 4 represents the utilities of GC. Generally, there 4 major points concerned with GC. The communication of green devices with cloud and virtualization are known to be as green computing communication (GCC). The overview of GCC is shown in Figure 5.

Control unit
Step 2: Decode instruction into command Step 3: Execute command Step 1: Fetch instruction the memory Step 4: Store instruction in memory

ALU
Control unit Main memory    [18]. An electronic CU has been developed for the control of the vehicle system by FPGA authors. The RISC processor (ARM) is used in combination with FPGA to perform parallel computing tasks [19]. A power-efficient CU on Virtex and the Spartan family FPGA was introduced to support the concepts of Green Communication researchers [20]. Authors have designed the integration of green communication on Virtex 4, Virtex 5, and Virtex 6 FPGA [21] in an energy-efficient instruction register. In [22], FPGA was used by the authors to produce a true random number by  3 Wireless Communications and Mobile Computing inducing metastability. In [23], photovoltaic simulation modules with FPGA were built in real time. In [24], researchers carried out a frequency change design of the arithmetic logic unit (ALU) for FPGA. Virtex-6 FPGA was used in [25] researchers to design a four-bit unregistered counter, allowing for clock and cutting. Random access memory (ROM) architecture for Virtex-6 FPGA was interfaced in [26]. In [27], researchers have used FPGA device to design a low power model for wireless data communication. In [28], researchers used energy-efficient techniques such as scaling the capacitance value of the capacitor of output load to design a green communication model of FIR Filter. With the help of Spartan-6 FPGA, authors have designed a green communication model of FIR Filter [29]. In [30], different families of FPGA devices have been used by the authors to develop a green UART for communication purpose. In [31], various FPGAs of the Spartan Group have been used for the implementation of energy-efficient transceiver model. In [32], researchers have developed a green CU with FPGA. For designing such model, authors have used HSTL and HSULIO standards. By using Pseudo Open Drain (POD) IO standards, an efficient FPGA model of ALU has been designed by the researchers [33]. To endorse green communication, researchers have designed an energyefficient model of instruction register on FPGA [21]. In [34], different FPGAs and SOC has been utilized to enhance the performance of FIR filter for data communication and communication channel. In [35], LVCMOS IO standards are considered to execute a power-efficient UART for green computing and green communication. In order to endorse the green wireless communication, authors have projected the idea of Vedic multiplier design on FPGA devices by       [36]. In [37], researchers built a powerefficient green communications paradigm employing the data outage and BUFG MB DSP state information (CSI) channel FPGA FF IO ports. In [38], authors have developed a green FF design for green wireless communication using FPGA architectures. In [39], researchers have used 28 nm FPGA device to design a thermal efficient as well as power-efficient CU to incorporate with green communication. Kintex Ultra-scale FPGA has been taken for modeling an energy-efficient CU for promoting the green communication [40,41]. Therefore, it has been observed that in the recent times, a lot of work has been done for incorporating the concepts of green communication and the energy/power efficient devices for future generations with the help of FPGAs, but a very few works have been done with respect to the implementation of the CU for green communication. Therefore, this work is all about the realization of CU on Zynq Ultra-Scale FPGA for promoting the values and ethics of green computing and green communication. The FPGA version of green computing model of CU is represented in Figure 6.

Experimental Setup
The ultrascale Zynq Soc FPGA board is used to set up the CU implementation. The VIVADO HLx architecture suite      Figure 11: Power calculation at 500 MHz.  Figure 8.

Results and Discussion
In addition, FPGA system dynamic power (DP) and static power (SP) are correlated with the power measurement of CUs using the Zynq FPGA [44]. The summation of both DP and SP is the overall total power (TP) consumption. The dynamic power is the device's leakage power release.

DP SP TP
The SP is the summation of I/O, logic (L/G), clock (CK), and signal (S/G). The power analysis of CU is done for five set of frequency value such as 100 MHz, 500 MHz, 1 GHz, 3 GHz, and 5 GHz, as shown in Figure 9.

Power Calculation for 100 MHz.
For the frequency of 100 MHz, the SP of the device is 0.589 W, which is 99% of the TP consumption. The SP is the summation of I/O, L/ G, CK, and S/G. Here, I/O power is 0.005 W, and the CK, L/G, and S/G power are less than 0.001 W. The DP, also called as leakage power, is 0.006 W, which is only 1% of the TP consumption. The TP for 100 MHz frequency is 0.595 W, as shown in Table 2 and Figure 10.

Power Calculation for 500 MHz.
For the frequency of 500 MHz, the TP consumption is 0.619 W, which is the summation of SP and DP which are 0.589 W and 0.030 W, respectively. The SP consumes 95% of the TP while DP consumes 5% of TP, as shown in Table 3 and Figure 11. 4.3. Power Calculation at 1 GHz. For the frequency of 1 GHz, the SP of the device is 0.590 W, which is 73% of the TP con-sumption. The SP is the summation of I/O, L/G, CK, and S/ G. Here, I/O power is 0.209 W, and the CK power is less than 0.001 W, while L/G and S/G power are 0.002 W and 0.003 W, respectively. The DP, also called as leakage power, is 0.214 W, which is 27% of the TP consumption. The TP for 1 GHz frequency is 0.804 W, as shown in Table 4 and Figure 12.   Figure 12: Power calculation at 1 GHz.  Figure 13: Power calculation at 3 GHz.  The SP consumes 76% of the TP while DP consumes 24% of TP, as shown in Table 5 and Figure 13.

Power Calculation at 5 GHz.
For the frequency of 5 GHz, the TP consumption is 0.893 W, which is the summation of SP and DP which are 0.590 W and 0.303 W, respectively. The SP consumes 66% of the TP while DP consumes 44% of TP, as shown in Table 6 and Figure 14.
Form the power calculation for different values of frequency, it is found to be that the power consumption is maximum for higher values of frequency, i.e., 5 GHz and minimum for 100 MHz. The TP consumed for all frequency values is depicted in Table 7 and Figure 15. It is also observed that there is an increase of 4.03% in TP as the frequency is raised to 500 MHz from 100 MHz. Also, the increment observed for 1 GHz, 3 GHz, and 5 GHz, which are 35.12%, 29.91%, and 50.08%, respectively.

Comparative Analysis
In this section, a comparison of TP consumption has been made with the existing works of CU on FPGA and with this work. In [18], with Spartan 6 FPPGA, the TP for CU is found to be 2.636 W, while in [20,40] Figure 14: Power calculation at 5 GHz.      [20,40], respectively. The TP consumption of CU with existing models and the proposed model is shown in Table 8 and Figure 16, respectively.

Conclusion and Future Scope
The transition to green communication is critical in this period, as energy crises can be seen all over the world. As a result of this study, several steps have been taken to promote the concepts of green communication and power-efficient devices. The implementation of CU is carried out on the Zynq SoC ultrascale FPGA, and the simulation of the CU circuit, resource usage, and power analysis is carried out on the VIVADO Hlx Design Suite. It has been found that as the clock frequency of the circuit is increased, the power consumption decreases. As a result, it can be inferred that the overall power consumption is reduced when the clock frequency is low. Therefore, it is observed that there is an increase of 4.03% in TP as the frequency is raised to 500 MHz from 100 MHz. Also, the increment observed for 1 GHz, 3 GHz, and 5 GHz, which are 35.12%, 29.91%, and 50.08%, respectively. Also, the TP of the proposed model is reduced by 77.42% from [18]. Similarly, the TP of this proposed model is reduced by 21.29% and 17.93% from [20,40], respectively, as shown in Figure 16. This CU circuit can be studied for other ultrascale and ultrascale plus FPGA devices in the future. Other power-saving methods, such as voltage, current, and capacitance scaling, can be used on the CU circuit as well. Impedance matching methods can also be used to make circuits more energy efficient with the aid of I/O specifications. For better performance, this FPGA design can later be converted to ASIC designs.

Data Availability
Data is available upon request.

Conflicts of Interest
The authors declare no conflicts of interest.