ABSTRACT
In this work, we present a methodology based on matrix factorization and gradient descent algorithm to predict the access points workload based on the users' behavior in Wi-Fi wireless networks, under academic and university environments. This knowledge is very useful when it is necessary to relocate determined access points due to the changing physical environment. We start collecting data about sessions and traffic of each user in each access point during certain time. From this information, prediction models are built using techniques that have been successfully applied to learning subjects, and next applied to cases where some access points should be reinforced because they support a high workload. For experimental purposes, we have collected data from the activity of 1,517 users in 37 access points of a wireless network in the School of Technology of the University of Extremadura, Spain, during October 2015. The obtained results are qualitatively valid with regard to the previous knowledge. The purpose of this methodology proposal is to apply the obtained results to improve the network infrastructure, as well as maintain the network against changing environments, since we can predict the access points workload when a determined device is removed, according to the usual activity of the network users.
- N. Banerjee and R. Kumar. Multiobjective network design for realistic traffic models. In Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation, pages 1904--1911, 2007. Google ScholarDigital Library
- S. Basagni, A. Carosi, and C. Petrioli. Heuristics for lifetime maximization in wireless sensor networks with multiple mobile sinks. In Proceedings of the 2009 IEEE International Conference on Communications, pages 1--6, 2009. Google ScholarDigital Library
- P. Boonma and J. Suzuki. Evolutionary constraint-based multiobjective adaptation for self-organizing wireless sensor networks. In Proc. of the 2nd ACM/IEEE/Create-Net/ICST International Conference on Bio-Inspired Models of Network, Information and Computing Systems (BIONETICS), pages 111--119, 2007.Google ScholarCross Ref
- L. Bottou. Large-scale machine learning with stochastic gradient descent. In Proceedings of 19th International Conference on Computational Statistics, pages 177--186, 2010.Google ScholarCross Ref
- S. Indu and A. Bhattacharyya. Self deployment of mobile sensor network for optimal coverage. International Journal of Engineering Science and Technology, 2(7):2968--2975, July 2010.Google Scholar
- D. Jannach, M. Z. amd A. Felfernig, and G. Friedrich. Recommender Systems. An Introduction. Cambridge University Press, 2011. Google ScholarDigital Library
- A. Konstantinidis and K. Yang. Multi-objective k-connected deployment and power assignment in wsns using a problem-specific constrained evolutionary algorithm based on decomposition. Computer Communications, 34:83--98, 2011. Google ScholarDigital Library
- Y. Koren. Factor in the neighbors: Scalable and accurate collaborative filtering. ACM Transactions on Knowledge Discovery from Data, 4:1--24, 2010. Google ScholarDigital Library
- Y. Koren, R. Bell, and C. Volinsky. Matrix factorization techniques for recommender systems. Computer, 42(8):30--37, 2009. Google ScholarDigital Library
- N. Thai-Nghe, L. Drumond, T. Horvath, A. Krohn-Grimberghe, A. Nanopoulos, and L. Schmidt-Thieme. Factorization techniques for predicting student performance. In Educational Recommender Systems and Technologies: Practices and Challenges, pages 129--153. IGI-Global, 2012.Google ScholarCross Ref
- M. Younis and K. Akkaya. Strategies and techniques for node placement in wireless sensor networks - a survey. Ad Hoc Networks, 6:621--655, 2008. Google ScholarDigital Library
Recommendations
Context-aware prediction of access points demand in Wi-Fi networks
We present a methodology based on matrix factorization and gradient descent to predict the number of sessions established in the access points of a Wi-Fi network according to the users behavior. As the network considered in this work is monitored and ...
Evaluating Next-Cell Predictors with Extensive Wi-Fi Mobility Data
Location is an important feature for many applications, and wireless networks may serve their clients better by anticipating client mobility. As a result, many location predictors have been proposed in the literature, though few have been evaluated with ...
Comments