Dr. R. Anil Kumar is currently working as an associate professor in the electronics and communication engineering department at Aditya College of Engineering Technology, Surampalem. He received a doctoral degree from JNTUK University, Kakinada. He published 15 technical papers in various international journals and presented six technical papers at various conferences. He is an expert in wireless communications, signal processing, and ML and DS. He published three patents. He is an associate member of IETE and ISTE.
EDUCATION
Ph.D awarded from JNTUK University Kannada
RESEARCH, TEACHING, or OTHER INTERESTS
Engineering, Computer Networks and Communications, Electrical and Electronic Engineering, Artificial Intelligence
Enhanced Sum Rate in IRS-THz System Using Hybrid Iterative Optimization Algorithm Fatima, Shahnaz, Anand, Vatsala, Thodupunoori, Padmavathi, Putchakayala, Venkateswara Rao, Relangi, Anil Kumar, et al. Passer Journal of Basic and Applied Sciences, 2025 This paper presents the use of intelligent reflecting surfaces (IRS) in terahertz (THz) communication systems. IRS is a promising technology to address challenges in THz communications such as severe path loss and limited scattering. However, the frequency-independent nature of IRS elements leads to beam squint effect in wideband THz systems. Mainly focusing on the impact of time delays. We propose a joint beamforming optimization framework to maximize the weighted Sum rate (WSR). This includes optimized base station beamforming, IRS reflection coefficients and time delays. The optimization problem is divided into subproblems, solved using techniques such as successive convex approximation (SCA), semidefinite relaxation (SDR) and fractional programming (FP). A Hybrid Iterative Optimization Algorithm iteratively achieves convergence. Simulation results show that our approach mitigates beam squint effects, enhancing system performance compared to conventional methods. The findings highlight the effectiveness of incorporating time delays into IRS design. This implementation offers a pathway to robust THz communication systems. From the simulation results, with 130 IRS elements, the Hybrid Scheme achieves a Sum rate of 55 bps/Hz. This is 37% higher than the SDR Scheme and 83% higher than the SCA Scheme. For bandwidth between 10 GHz and 20 GHz, the hybrid Scheme maintains a stable performance of around 700 bps/Hz. It is over 2.3 times higher than the SDR Scheme and nearly 3.5 times higher than the SCA Scheme.
Optimizing MIMO-GFDM for 5G and Beyond: A Joint Approach With Subcarrier Index Modulation and Constellation Precoding R. Anil Kumar, Vivek Rajpoot International Journal of Communication Systems, 2025 In the future, massive growth in mobile users in various applications is anticipated. To cope with this growth, the enhanced radio access technology (eRAT) can provide a solution to improve diversity gain and spectral efficiency. The state‐of‐the‐art generalized frequency division multiplexing (GFDM) modulation techniques are combined with a multi‐input multi‐output (MIMO) antenna system to achieve better resource allocation, out‐of‐band emission, and signal strengths over the existing orthogonal multiple access (OMA) techniques. In the proposed research paper, the subcarrier index modulation (SIM) and constellation precoding (CP) techniques are combined with the MIMO‐GFDM system to improve the performance of diversity gain, spectral efficiency, and energy efficiency. The system is called the SIM‐CP‐MIMO‐GFDM. In the SIM‐CP‐MIMO‐GFDM system, first, a few subcarriers are activated by index modulation bits in quadrature/in‐phase dimensions, and then data symbols are constellation precoded. At the receiving end, the data blocks are detected using QR decomposition‐based maximum likelihood (ML) detection. Finally, the paper incorporates theoretical and simulation result analysis and compares the proposed SIM‐CP‐MIMO‐GFDM system with existing systems. The proposed SIM‐CP‐MIMO‐GFDM outperforms the existing systems.
Optimization of Deep Learning Technique for OFDM Receivers in 6G Wireless Communications Kasetty Lakshmi Narasimha, V. Saraswathi, Mummidi SubbaRaju, M. Koteswara Rao, Kapula Kalyani Kalyani, Anil Kumar R Eai Endorsed Transactions on Internet of Things, 2025 INTRODUCTION: This paper presents an innovative deep learning-based optimization technique for orthogonal frequency division multiplexing (OFDM) receivers in wireless communication systems.OBJECTIVES: The proposed method utilizes an enhanced deep convolutional neural network (Enhanced DCNN) architecture with a time-frequency domain fusion mechanism to address the issues of interference and temporal information loss. The model incorporates attention mechanisms and causal convolutions to extract long-term dependencies within the received OFDM signals. It enables accurate channel estimation and signal recovery.METHODS: The methodology is validated using simulations based on 3GPP-defined channel models. It includes extended typical U (ETU), extended pedestrian A (EPA) and extended vehicular A (EVA) across varying signal-to-noise ratio (SNR) conditions.RESULTS: Results demonstrate that the proposed receiver significantly improves bit error rate (BER) performance compared to traditional Least Squares (LS) and LMMSE methods. Particularly, in scenarios with large delay spreads and high mobility. Additionally, the model has a lower computational complexity (CC) and thus is appropriate for real-time implementation.CONCLUSION: We view this work as a strong scheme to improve the performance of OFDM systems in future wireless networks.
Performance Analysis of Blockage Detection in 6G Wireless Networks Naziya Hussain, Vatsala Anand, Anuragh Vijjapu, G.A. Arun Kumar, R. Anil Kumar, V. Preethi International Journal of Electrical and Electronics Research, 2025 This paper presents a proactive blockage detection algorithm and performance for sub-terahertz (sub-THz) communication systems. It is a critical technology for next-generation wireless networks. Human body blockage is a significant challenge for millimeter wave and sub-THz systems. It often causes severe signal degradation and connectivity loss. Current solutions, such as time-domain approaches or machine learning models. These methods suffer from high computational complexity, limited accuracy and an inability to detect blockages before they occur. The proposed algorithm overcomes these limitations by applying spectral-domain analysis using the short-time Fourier transform (STFT). It provides precise detection of early blockage conditions. The blockage detection algorithm adopts a simple and threshold-based detection mechanism. The algorithm operates in three stages: initialization, active monitoring, and blockage recovery. The proposed algorithm demonstrates a detection probability of 81% and highlights its effectiveness in reliable detection. It detects blockages at least before their occurrence. Additionally, the algorithm achieves a higher mean time to blockage of up to 150ms under optimal parameter settings. These findings highlight the algorithm's effectiveness in mitigating connectivity issues in dense indoor deployments. It is adaptable to different network configurations and is suitable for various deployment scenarios. The proposed solution is well-suited for ultra-reliable low-latency communications, augmented reality, autonomous vehicles, and industrial IoT. These outcomes mark a key step toward seamless 6G communication.
PAPR reduction of GFDM signals using moving average filtering scheme for future wireless communications R. Anil KUMAR Sigma Journal of Engineering and Natural Sciences, 2025 Generalized Frequency Division Multiplexing (GFDM) is a new method for transmitting data in blocks.It is being considered for 5G because it offers better spectral efficiency, lower latency, and more scalability than Orthogonal Frequency Division Multiplexing (OFDM).However, GFDM still faces a problem with high Peak-to-Average Power Ratio (PAPR), which can reduce the system's performance.Therefore, we proposed Moving Average Filtering (MAF) scheme to reduce the high PAPR of the GFDM signals.LAMF operates by taking an average number of GFDM-modulated signal sample points to produce the smoothed output sample points.The computed smoothed values reduce the random high peak noise samples that depend on the length of the filter.In this article, we presented the mathematical analysis of the PAPR of the GFDM signal and the design of a linear average moving filter.The proposed scheme's PAPR and symbol error rate (SER) is compared with different filter lengths and traditional techniques.The proposed scheme's PAPR and Symbol Error Rate (SER) are compared with different filter lengths and traditional techniques.Our findings demonstrate that the PAPR of the MAF_GFDM signal is reduced to 3.22 dB at CCDF = 10 -3 with a filter length L=64, while the symbol error rate is improved to 0.510 -4 .Additionally, the minimum mean square error estimation method is identified as the best estimator, achieving a very low symbol error rate (0.001) at a signal-to-noise ratio of 11 dB.These simulation results obtained using MATLAB demonstrate the effectiveness of the proposed scheme in significantly reducing PAPR and improving SER.
Optimized Task Offloading in D2D-Assisted Cloud-Edge Networks Using Hybrid Deep Reinforcement Learning Navya Kailasam, Srilatha Yalamati, V. S. N. Murthy, Venkateswara Rao P, R. Anil Kumar, K. Jayaram Kumar International Journal of Basic and Applied Sciences, 2025 The modern communication network depends highly on Device-to-Device (D2D) technology as an essential foundation. Direct communication allows devices to exchange information among themselves. Cloud-edge-device networks enable tasks to execute through several operational procedures. A device working at capacity executes local tasks or transfers them directly to an inactive device by means of D2D technology. The device has two options for delivering workloads, namely an edge-server transfer or a direct cloud-server transfer. Existing methods do not fully exploit D2D-assisted offloading. Such systems fail to maximize the benefits that stem from combining cloud-edge-device op-operations. This makes resource distribution a complex challenge that needs an optimized solution. Traditional solutions find it difficult to produce efficient system solutions. The presented work describes an approach for task offloading mechanisms. The technique determines overall system expenses through optimized management of time, together with energy usage. The method operates to optimize all four critical system factors: task selection and transmission power, with rate and computational resource distribution. The proposal utilizes a combina-tion of deep reinforcement learning methods through SD3. The proposed method merges Softmax Deep Double Deterministic Policy Gradients (SD3) with numerical techniques to achieve its operations. The proposed method operates on multiple smaller components of the primary issue. The SD3-based DRL method controls offloading decisions throughout the system, and the numerical techniques manage power and resource allocation. Extensive simulations were conducted. Seven different scenarios were tested. Research compared the pro-posed method against four traditional solution approaches. Research findings demonstrate the superiority of the proposed solution. The technique both lessens system expenses and optimizes resource usage while generating better operational efficiency. A novel hybrid DRL-based approach for task offloading constitutes the main contribution of this work. The system improves cloud-edge-device partnerships by enabling D2D communication. Machine learning unions with numerical methods create an effective strategy to solve complex optimization tasks.
Blockchain-Based Edge Computing: Joint Task Offloading And Mining With Multi-Agent Reinforcement Learning Journal of Theoretical and Applied Information Technology, 2025
