International Journal of Computer Science and Technology

International Journal of Computer Science and Technology | Vol. 4, No. 12, December 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000227

Article Type: Review Paper

Title: Neuromorphic Computing Architecture Paradigms and Silicon Implementations: A Comparative Analysis

Names of Authors: Kofi Mensah¹, Yuki Tanaka²

Authors’ Affiliations: ¹Department of Computer Engineering, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana; ²Advanced VLSI Systems Laboratory, Tokyo Tech, Tokyo, Japan

Abstract: Von Neumann processing architectures encounter severe energy bottlenecks and memory speed limitations when running modern deep learning model structures. Neuromorphic computing presents an alternative framework by designing silicon architectures that mimic biological neural networks. This review paper provides a comparative analysis of modern neuromorphic silicon implementations, including Intel Loihi, IBM TrueNorth, and SpiNNaker. We evaluate each hardware platform across operational vectors such as synaptic density, energy dissipation, and on-chip learning efficiency. The study aggregates performance datasets compiled from 40 experimental research papers published over the past six years. Our findings show that neuromorphic designs cut power usage by up to 1,000x compared to classic GPU architectures when processing sparse temporal streams. However, configuring non-von Neumann systems requires specialized programming tools and remains limited by hardware constraints during dense matrix calculations. This paper presents a taxonomical framework to guide hardware designers in matching neuromorphic platforms with specific edge applications.

Keywords: Neuromorphic Computing, Non-Von Neumann Architecture, Spiking Neural Networks, Silicon Implementations, Energy Efficiency, VLSI

Manuscript Timeline: Received: March 20, 2023; Revised: May 12, 2023; Accepted: June 28, 2023; Published: December 10, 2023

Citation: Mensah, K., & Tanaka, Y. (2023). Neuromorphic Computing Architecture Paradigms and Silicon Implementations: A Comparative Analysis. International Journal of Computer Science and Technology, 4(12), 1–8. DOI: 10.46882/2023/IJCST/000227

International Journal of Computer Science and Technology | Vol. 4, No. 11, November 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000226

Article Type: Original Research Paper

Title: Context-Aware Document Summarization Architecture Using Transformer-Based Attention Weighting

Names of Authors: Anna Kowalski¹, Hans-Jürgen Integrated²

Authors’ Affiliations: ¹Faculty of Electronics and Information Technology, Warsaw University of Technology, Warsaw, Poland; ²Institute for Cognitive Systems Research, University of Stuttgart, Stuttgart, Germany

Abstract: Automated document summarization frameworks frequently generate disconnected output strings or omit critical context when processing long technical texts. This paper introduces an abstractive document summarization model that uses a context-aware transformer architecture. The framework features a targeted attention-weighting layer that highlights domain-specific technical terminology and structural document headings. This layer prevents the model from dropping high-impact semantic connections across paragraphs. We integrated a pointer-generator network layer to copy unique named entities and numerical metrics accurately from source text fields into the summaries. The neural model was trained and validated using standard CNN/DailyMail and arXiv technical text corpora. The validation results show a ROUGE-L score improvement of 4.35 points over baseline text-summarization models. Human evaluation tracking confirmed that the generated summaries maintained high context consistency, minimizing factual hallucination rates to less than 1.8% across tested articles.

Keywords: Natural Language Processing, Document Summarization, Transformers, Attention Mechanism, Factual Consistency, Text Mining

Manuscript Timeline: Received: March 15, 2023; Revised: May 08, 2023; Accepted: June 22, 2023; Published: November 15, 2023

Citation: Kowalski, A., & Integrated, H. -J. (2023). Context-Aware Document Summarization Architecture Using Transformer-Based Attention Weighting. International Journal of Computer Science and Technology, 4(11), 1–8. DOI: 10.46882/2023/IJCST/000226

International Journal of Computer Science and Technology | Vol. 4, No. 10, October 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000225

Article Type: Original Research Paper

Title: Optimizing Spatial Query Performance in Distributed GIS Architectures via Multi-Level Grid Indexing

Names of Authors: Min-Ji Kim¹, David Miller²

Authors’ Affiliations: ¹Spatial Information Systems Lab, Pusan National University, Busan, South Korea; ²Geospatial Technologies Group, TechCorp Innovations, San Francisco, USA

Abstract: Distributed Geographic Information Systems (GIS) face major performance bottlenecks when executing complex spatial queries across multi-terabyte environmental datasets. Conventional indexing methods like standard R-trees encounter severe overlap issues when processing high-density urban coordinates, resulting in long query lookup delays. This paper proposes a multi-level grid indexing framework optimized for distributed spatial databases. The framework divides geographic coordinate spaces into hierarchical, non-overlapping grid cells, partitioning geospatial datasets balanced across cluster nodes. An adaptive spatial data placement algorithm balances workloads among servers based on historical query access frequencies. We validated this indexing architecture across a 16-node distributed cluster using a 500 GB global transport dataset. The experimental findings indicate a 44.1% reduction in total spatial query execution time compared to traditional distributed R-tree implementations. The cluster achieved a sustained query processing capacity of 12,400 queries per second, demonstrating its scalability for web-mapping infrastructures.

Keywords: Geographic Information Systems, Distributed Indexing, Spatial Queries, Grid Partitioning, Workload Balancing, Database Performance

Manuscript Timeline: Received: March 10, 2023; Revised: May 02, 2023; Accepted: June 15, 2023; Published: October 12, 2023

Citation: Kim, M. -J., & Miller, D. (2023). Optimizing Spatial Query Performance in Distributed GIS Architectures via Multi-Level Grid Indexing. International Journal of Computer Science and Technology, 4(10), 1–8. DOI: 10.46882/2023/IJCST/000225

International Journal of Computer Science and Technology | Vol. 4, No. 9, September 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000224

Article Type: Original Research Paper

Title: Resilient Multi-UAV Swarm Coordination under Targeted Communication Jamming Conditions

Names of Authors: Carlos Silva¹, Amara Okafor²

Authors’ Affiliations: ¹Department of Systems Engineering, Technological Institute of Aeronautics, São José dos Campos, Brazil; ²Department of Computer Science, University of Ibadan, Ibadan, Nigeria

Abstract: Unmanned Aerial Vehicle (UAV) swarms require continuous peer-to-peer data coordination to execute spatial surveying operations safely and avoid mid-air collisions. However, targeted communication jamming attacks can sever inter-node links, leading to swarm fragmentation or mission failure. This study presents a resilient decentralized swarm coordination protocol developed to handle localized communication dropouts. The framework introduces a dynamic potential-field routing topology that recalculates vehicle trajectories using onboard optical flow data when communication signals are lost. This allows independent aerial units to infer neighboring maneuvers based on visual observation flags rather than relying on wireless data feeds. We tested this tracking mechanism within a simulated 3D operational space using 24 moving aerial vehicles under active signal jamming variables. The simulation results show that the swarm maintained a 91.6% structural cohesion level during active signal dropouts. The system completely avoided inter-vehicle collisions while restricting tracking trajectory errors to ±1.4 m during autonomous recovery phases.

Keywords: UAV Swarms, Communication Jamming, Decentralized Coordination, Optical Flow, Collision Avoidance, Network Resilience

Manuscript Timeline: Received: March 01, 2023; Revised: April 20, 2023; Accepted: May 29, 2023; Published: September 08, 2023

Citation: Silva, C., & Okafor, A. (2023). Resilient Multi-UAV Swarm Coordination under Targeted Communication Jamming Conditions. International Journal of Computer Science and Technology, 4(9), 1–8. DOI: 10.46882/2023/IJCST/000224

International Journal of Computer Science and Technology | Vol. 4, No. 8, August 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000223

Article Type: Original Research Paper

Title: Optimizing Deep Neural Network Inference on Edge TPU Devices through Integer Quantization Protocols

Names of Authors: Amina Al-Mansoor¹, Dieter Schmidt²

Authors’ Affiliations: ¹Department of Computer Engineering, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia; ²Institute for Computer Architecture, Technical University of Munich, Munich, Germany

Abstract: Running complex object detection networks on localized edge hardware remains limited by the high memory requirements and power consumption profiles of floating-point processing layers. This research presents an integer quantization protocol designed to compress deep learning networks for edge-layer Tensor Processing Unit (TPU) deployments. Our methodology implements an optimized post-training quantization algorithm that maps 32-bit floating-point parameters (FP32) into 8-bit integer formats (INT8). The system uses a dynamic scaling factor calculation across intermediate network layers to minimize quantization noise accumulation. We evaluated this compression architecture using MobileNet-V3 and YOLO-V5 models running on embedded edge TPU modules. The experimental measurements indicate a 72.3% reduction in the total memory space required by the neural model. The total frame inference rate increased by 2.8x, achieving a processing latency of 14 ms per image frame. Crucially, the overall classification accuracy degraded by only 0.42% compared to the original uncompressed model framework.

Keywords: Edge TPU, Model Compression, Integer Quantization, Deep Neural Networks, Computer Vision, Resource Optimization

Manuscript Timeline: Received: February 22, 2023; Revised: April 15, 2023; Accepted: May 25, 2023; Published: August 14, 2023

Citation: Al-Mansoor, A., & Schmidt, D. (2023). Optimizing Deep Neural Network Inference on Edge TPU Devices through Integer Quantization Protocols. International Journal of Computer Science and Technology, 4(8), 1–8. DOI: 10.46882/2023/IJCST/000223

International Journal of Computer Science and Technology | Vol. 4, No. 7, July 2023 | pp. 1–8

DOI: 10.46882/2023/IJCST/000222

Article Type: Original Research Paper

Title: Quantifying Malicious Smart Contract Behavior via Automated Abstract Syntax Tree Feature Mapping

Names of Authors: Sanjay Nair¹, Takashi Sato²

Authors’ Affiliations: ¹Cyber Security Research Center, Indian Institute of Information Technology, Bangalore, India; ²School of Engineering, Tokyo Institute of Technology, Tokyo, Japan

Abstract: Decentralized application protocols running on public blockchain ecosystems remain highly vulnerable to smart contract vulnerabilities, exposing digital assets to targeted reentrancy and integer overflow exploits. Manual source-code validation is slow and cannot scale with the rapid deployment rate of smart contracts. This article introduces an automated vulnerability identification pipeline that uses structural feature mapping of Abstract Syntax Trees (ASTs). The pipeline extracts syntactic dependencies and data-flow pathways from smart contract source files, converting them into structured low-dimensional vector representations. A random forest classifier evaluates these feature maps to isolate architectural flaws before compilation or deployment. We tested this validation framework using a verification dataset of 4,500 checked Ethereum smart contracts. The analysis results prove that the tool achieves a true positive classification rate of 97.4%, surpassing standard rule-based pattern matching tools by 14.5%. The average analysis duration recorded was 240 ms per contract file, making it suitable for continuous integration (CI/CD) pipelines.

Keywords: Blockchain, Smart Contracts, Vulnerability Assessment, Abstract Syntax Tree, Automated Detection, Cybersecurity

Manuscript Timeline: Received: February 18, 2023; Revised: April 10, 2023; Accepted: May 20, 2023; Published: July 11, 2023

Citation: Nair, S., & Sato, T. (2023). Quantifying Malicious Smart Contract Behavior via Automated Abstract Syntax Tree Feature Mapping. International Journal of Computer Science and Technology, 4(7), 1–8. DOI: 10.46882/2023/IJCST/000222