Ahmed Hamdi Sakr

Dr. Ahmed Hamdi Sakr


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 February 2022 in IEEE Transactions on Intelligent Vehicles
Verified
Journal Article
Closed access
HWHao M. Wang
Hao M. Wang
SASergei S. Avedisov
Sergei S. Avedisov
TMTamás G. Molnár
Tamás G. Molnár
ASAhmed Hamdi Sakr
Ahmed Hamdi Sakr
OAOnur Altintas
Onur Altintas
GOGábor Orosz
Gábor Orosz
In this paper, we establish the concept of conflict analysis and demonstrate its applicability to aid the decision making of vehicles at different levels of automation and cooperation. In particular, we assume that the participating vehicles are equipped with vehicle-to-everything (V2X) communication and study cooperative maneuvers under status sharing and intent sharing. Conflict-free maneuvering strategies are developed and communication requirements for such strategies are determined. To demonstrate the developed framework, we investigate a scenario with an ego vehicle merging to a main road while another vehicle is approaching on that road. The results are demonstrated by both experimental data using real vehicles on a test track and simulations based on real highway data.

18 November 2024 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
AUAshab Uddin
Ashab Uddin
ASAhmed Hamdi Sakr
Ahmed Hamdi Sakr
NZNing Zhang
Ning Zhang
Vehicular edge computing enables real-time decision-making by offloading vehicular computation tasks to edge servers along roadways. This paper focuses on optimizing offloading and scheduling these tasks, with an emphasis on task prioritization to maximize task completion within deadlines while minimizing latency and energy consumption across all priority levels. We propose a prioritized Deep Q-Network (DQNP) that optimizes long-term rewards through a priority-scaled reward system for each priority level, guiding the deep reinforcement learning (DRL) agent to select optimal actions. The model dynamically adjusts task selection based on environmental conditions, such as prioritizing tasks with higher deadlines in poor channel states, ensuring balanced and efficient offloading across all priority levels. Simulation results demonstrate that DQNP outperforms existing baseline algorithms, increasing task completion by 14%, particularly for high-priority tasks, while reducing energy consumption by 8% and maintaining similar latency. Additionally, the model mitigates resource starvation for lower-priority tasks, achieving task selection rates of 27%, 32%, and 42% for low-, medium-, and high-priority tasks, with completion ratios of 88%, 87%, and 86%, respectively, reflecting balanced resource allocation across priority classes.

22 June 2021 in IEEE Transactions on Wireless Communications
Verified
Journal Article
Closed access
JJJeya Pradha Jeyaraj
Jeya Pradha Jeyaraj
Martin HaenggiMartin Haenggi
Martin Haenggi
ASAhmed Hamdi Sakr
Ahmed Hamdi Sakr
HLHongsheng Lu
Hongsheng Lu
A comprehensive vehicular network analysis requires modeling the street system and vehicle locations. Even when Poisson point processes (PPPs) are used to model the vehicle locations on each street, the analysis is barely tractable. That holds for even a simple average-based performance metric—the success probability, which is a special case of the fine-grained metric, the meta distribution (MD) of the signal-to-interference ratio (SIR). To address this issue, we propose the transdimensional approach as an alternative. Here, the union of 1D PPPs on the streets is simplified to the transdimensional PPP (TPPP), a superposition of 1D and 2D PPPs. The TPPP includes the 1D PPPs on the streets passing through the receiving vehicle and models the remaining vehicles as a 2D PPP ignoring their street geometry. Through the SIR MD analysis, we show that the TPPP provides good approximations to the more cumbrous models with streets characterized by Poisson line/stick processes; and we prove that the accuracy of the TPPP further improves under shadowing. Lastly, we use the MD results to control network congestion by adjusting the transmit rate while maintaining a target fraction of reliable links. A key insight is that the success probability is an inadequate measure of congestion as it does not capture the reliabilities of the individual links.

26 May 2021 in IEEE Vehicular Technology Magazine
Verified
Journal Article
Closed access
SASergei Avedisov
Sergei Avedisov
ASAhmed Hamdi Sakr
Ahmed Hamdi Sakr
THTakamasa Higuchi
Takamasa Higuchi
AYAkihiko Yamamuro
Akihiko Yamamuro
OAOnur Altintas
Onur Altintas

20 September 2024 in IEEE Access
Verified
Journal Article
Open access
ASAhmed Hamdi Sakr
Ahmed Hamdi Sakr
Collective perception enables connected vehicles to share detailed environmental data, significantly enhancing situational awareness and safety. This data sharing is crucial for the functioning of modern vehicular networks, but it introduces the challenge of managing redundant information, which can congest communication channels and degrade network performance. To address this challenge, several redundancy mitigation rules have been proposed and extensively evaluated to filter out unnecessary data. This work investigates the impact of different redundancy mitigation rules on the performance of connected vehicular networks with collective perception under different market penetration rates. Additionally, the study introduces a set of hybrid rules designed to optimize this balance for collective perception services in vehicular networks. These hybrid rules are compared to scenarios without object filtering and other existing redundancy mitigation rules. Key performance metrics include channel busy ratio, environment awareness ratio, redundancy level, and the age of information. By analyzing the metrics as a function of the distance between the reported object and the receiving connected vehicle, the study identifies key trends in balancing redundancy reduction with information freshness under diverse network conditions. The results demonstrate that hybrid redundancy mitigation rules outperform existing approaches by effectively balancing channel load, redundancy level, and environment awareness, while maintaining lower age of information values. This balance is particularly crucial for safety-critical objects in close proximity to the connected vehicle. The findings highlight the importance of intelligent redundancy mitigation strategies in enhancing the timeliness and reliability of information in densely populated vehicular networks, ensuring the efficient and safe operation of connected vehicles.

10 December 2025 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
AUAshab Uddin
Ashab Uddin
ASAhmed Hamdi Sakr
Ahmed Hamdi Sakr
NZNing Zhang
Ning Zhang
This paper addresses the problem of task offloading and scheduling in Vehicular Edge Computing (VEC) by incorporating dynamic task prioritization across distributed zones. The objective is to maximize task completion within deadline constraints while minimizing latency and energy consumption. To this end, the problem is formulated as a Partially Ob servable Markov Decision Process (POMDP), and solved using a Counterfactual Multi-Agent (COMA) reinforcement learning framework with centralized training and decentralized execution. The proposed approach enables distributed agents to jointly determine task priority classes, edge server associations, and CPU frequencies, leveraging counterfactual credit assignment to coordinate decision-making in both cooperative and competitive environments. Simulation results demonstrate that COMA achieves zone-aware task prioritization with improved fairness and resource distribution. It matches or exceeds the performance of centralized value-based baselines and reduces average latency by approximately 6% compared to centralized Deep Q-Learning. Additionally, it offers more balanced prioritization than multi agent deterministic policy gradient methods, highlighting its effectiveness in dynamic, distributed edge environments.