arXiv subjects

Rajib Mall

Publications and source records attributed to Rajib Mall.

4 recordsLinked to original sources

arXiv2026

RL+AHP: A Novel Reinforcement Learning driven AHP for Slice Aware mode selection in D2D enabled Heterogeneous Networks

The mode selection problem in device-to-device communication (D2D) enabled Fifth generation (5G) heterogeneous networks (HetNet) aims prioritizing four key performance indicators (KPIs) namely data rate, latency, reliability and jitter across three slices: enhanced mobile broadband (eMBB), ultra reliable low latency (uRLLc) and massive machine type communications (mMTC). Such priority assignment must be \emph{traded off} among three access technologies, i.e., Long Term Evolution advanced (LTE-A), New Radio (NR) and D2D, while minimizing handover frequency. In existing mode selection approaches for HetNet, slice specific quality of service (QoS) requirements are largely ignored. In this work, a novel mode selection algorithm is proposed by combining a two level Analytic Hierarchy Process (AHP) with a Reinforcement Learning (RL) method. While the two level AHP facilitates decision making based on multiple criteria (i.e., KPIs) and options (i.e., LTE-A, NR, D2D mode), the RL approach computes the weights of each criteria based on the feedback from the environment. Simulation results show that our proposed algorithm outperforms related works in terms of the major KPIs for all three slices. For eMBB applications, our approach increases throughput by $33\%$; for uRLLc applications, our approach significantly decreases latency and BER ($27\%$ and $10\%$ respectively) and for mMTc applications, our approach significantly decreases latency ($44\%$). Moreover, it has been shown that the proposed RL+AHP approach outperforms the existing DRL based approaches in terms of CPU usage when the number of criteria is reasonably low ($<6$).

cs.NI↗

arXiv2025

AI-Driven Early Warning Systems for Student Success: Discovering Static Feature Dominance in Temporal Prediction Models

Early identification of at-risk students is critical for effective intervention in online learning environments. This study extends temporal prediction analysis to Week 20 (50% of course duration), comparing Decision Tree and Long Short- Term Memory (LSTM) models across six temporal snapshots. Our analysis reveals that different performance metrics matter at different intervention stages: high recall is critical for early intervention (Weeks 2-4), while balanced precision-recall is important for mid-course resource allocation (Weeks 8-16), and high precision becomes paramount in later stages (Week 20). We demonstrate that static demographic features dominate predictions (68% importance), enabling assessment-free early prediction. The LSTM model achieves 97% recall at Week 2, making it ideal for early intervention, while Decision Tree provides stable balanced performance (78% accuracy) during mid-course. By Week 20, both models converge to similar recall (68%), but LSTM achieves higher precision (90% vs 86%). Our findings also suggest that model selection should depend on intervention timing, and that early signals (Weeks 2-4) are sufficient for reliable initial prediction using primarily demographic and pre-enrollment information.

cs.LG↗

arXiv2024

Effective Fault Localization using Probabilistic and Grouping Approach

Context: Fault localization (FL) is the key activity while debugging a program. Any improvement to this activity leads to significant improvement in total software development cost. There is an internal linkage between the program spectrum and test execution result. Conditional probability in statistics captures the probability of occurring one event in relationship to one or more other events. Objectives: The aim of this paper is to use the conception of conditional probability to design an effective fault localization technique. Methods: In the paper, we present a fault localization technique that derives the association between statement coverage information and test case execution result using condition probability statistics. This association with the failed test case result shows the fault containing the probability of that specific statement. Subsequently, we use a grouping method to refine the obtained statement ranking sequence for better fault localization. Results: We evaluated the effectiveness of proposed method over eleven open-source data sets. Our obtained results show that on average, the proposed CGFL method is 24.56% more effective than other contemporary fault localization methods such as D*, Tarantula, Ochiai, Crosstab, BPNN, RBFNN, DNN, and CNN. Conclusion: We devised an effective fault localization technique by combining the conditional probabilistic method with failed test case execution-based approach. Our experimental evaluation shows our proposed method outperforms the existing fault localization techniques.

cs.SE↗

arXiv2011

MiCi: A Novel Micro-Level Temporal Channel Imploration for Mobile Hosts

The exponential increase of multimedia services by the mobile users requires seamless connectivity with cost effective Quality of Service QoS provisioning. For providing such on-demand QoS, the network needs to utilize the radio channels among the Mobile Hosts (MHs) effectively. We use vector genetic algorithm VGA for temporal imploration of sharable channel(s) from the neighbouring cells to fulfill the needs of a cell. We propose a new micro-level temporal channel imploration mechanism MiCi, which promptly allocates available borrowing channel s of the neighbouring cell(s) to the needy cell. The novelty of MiCi is scalability, high availability, and on demand allocation of the channels to the desired cells. The performance of our model has been tested by simulation against a standard FCA scheme as well as a Greedy Borrowing Heuristic. In all the test cases MiCi shows promising results in comparison to both the schemes.

cs.NI↗