Author(s)

HongLing Zhang

Affiliation(s)

School of Information Engineering, Zhongyuan Institute Of Science And Technology, Xuchang 461113, Henan, China.

Corresponding Author

HongLing Zhang

ABSTRACT

In recent years, Radio Frequency Identification (RFID) technology has seen expanding applications across both industrial production and daily life, driving growing demand for efficient tag reading systems.When faced with a large numeral of tags, the Dynamic Framed Slotted ALOHA (DFSA ) algorithm keeps the throughput at a high position for most of the time. The chief principle is to dynamically adjust the frame length based on the response consequence of the triumphant transmission of the earlier frame, so that during the data container transmission process, the length of each frame is close to the number of data packets that need to be transmitted within the range that the information transmission system can receive. Based on the analysis of traditional algorithms, this paper proposes a new tag number estimation algorithm. This algorithm based on Transformer, residual connections, and Multi-Layer Perceptron (MLP), combined with algorithms such as tag grouping techniques. Compared with traditional algorithms, the algorithm proposed in this paper addresses the shortcomings of the traditional ALOHA protocol in dynamic frame slot adjustment, collision avoidance, and throughput optimization by introducing deep learning. It significantly improves the effectiveness and reliability of RFID systems and is able of maintaining a high data storage rate even in scenarios with large amounts of data.

KEYWORDS

RFID; DFSA; MLP-Transformer

CITE THIS PAPER

HongLing Zhang. Deep learning-enhanced dynamic frame slotted ALOHA optimization algorithm. World Journal of Information Technology. 2025, 3(2): 68-74. DOI: https://doi.org/10.61784/wjit3036.

REFERENCES

[1] Maksimenko A, Dobrykh D, Yusupov I, et al. Miniaturization limits of ceramic UHF RFID tags. Scientific Reports, 2025, 15(1): 10984.

[2] Masekwana F, Jokonya O. Factors affecting the adoption of RFID in the food supply chain: A Systematic Literature Review. Frontiers in Sustainable Food Systems, 2025, 8: 1497585.

[3] Claucherty E, Cummins D, Aliakbarian B. RFID Unpacked: A Case Study in Employing RFID Tags from Item to Pallet Level. Electronics, 2025, 14(2): 278.

[4] Wu Y, Lin J, Chen H, et al. A transformer-based double-order RFID indoor positioning system. Expert Systems with Applications, 2025: 126530.

[5] Maimouni M, Abou El Majd B, Bouya M. Optimising RFID network planning problem using an improved automated approach inspired by artificial neural networks. Information Sciences, 2025: 121927.

[6] Lasantha L, Ray B, Karmakar N. Trade-Off Analysis for Array Configurations of Chipless RFID Sensor Tag Designs. Sensors, 2025, 25(6): 1653.

[7] Wang Z, Mao S. Generative AI-Empowered RFID Sensing for 3D Human Pose Augmentation and Completion. IEEE Open Journal of the Communications Society, 2025.

[8] Lazaro A, Cujilema M R, Villarino R, et al. A novel approach for wine anti-counterfeiting using laser-induced graphene chipless RFID tags on cork. Scientific Reports, 2025, 15(1): 12750.

[9] Liu H, Meng Z, Li C, et al. Modeling for Phase Decoupling to Detect the Orientation and Position of Moving Objects with Simple RFID Array. IEEE Transactions on Instrumentation and Measurement, 2025.

[10] Ming Dongyue, Wang Shangpeng, Lei Ming, et al. Improved Dynamic Framed Slotted ALOHA Algorithm Combined with BP Neural Network . Mini-Micro Systems, 2021, 42(09): 1920-1923.

[11] Li Z, Li Z. An RFID anti-collision algorithm integrated with LSTM. 2024 IEEE 4th International Conference on Power, Electronics and Computer Applications (ICPECA), IEEE, 2024: 985-989.

[12] Zhou Q. ALOHA Improvement Algorithm for Dynamic Frame Time Slots with Deep Learning. 2024 IEEE 4th International Conference on Data Science and Computer Application (ICDSCA), IEEE, 2024: 671-676.