The wide flexibility of LTE resource allocation scheme has led to the definition of various schedulers that attempt to maximize the quality of the service offered to the different users, depending on their channel conditions. Unfortunately, providing service guarantees in dynamic channel conditions typically requires a cost in terms of spectral efficiency of the transmission resource allocation. In this work, we investigate this tradeoff and propose a resource allocation scheme that adapts the service level guarantees to the average channel conditions of the users, in order to provide fair resource allocation among users with homogeneous channel conditions, while improving the cell spectral efficiency.

A performance analysis is carried out by comparing the proposed scheduler with other well known schedulers. Results show that the proposed method can improve the cell capacity, while guaranteeing long-term throughput fairness among users of the same class. In addition, we analyze the short-term throughput provided by the proposed scheduler and provide a semi-analytical model to assess the gap with respect to the long-term performance.