This paper studies the potential benefit of joint load balancing and interference mitigation in the uplink wireless network through optimized cell association and user scheduling across multiple cells. Coordinating uplink transmission is considerably more challenging than the downlink, because the uplink interference strongly depends on the transmission patterns of other users in the neighboring cells. This paper formulates the uplink joint cell association and scheduling problem as a network utility maximization problem. We propose a series of problem reformulations based fractional programming, and recursively solve the resulting convex approximation to reach a local optimum of the proportionally fair resource allocation problem. Numerical results suggest significant performance improvement for the proposed method as compared to the benchmarks.