The huge growth in demand for wireless data, combined with shortages of spectrum, has led to an urgent need for spectral efficiency and cell-edge performance improvements in cellular networks. Macrocells are shrinking in size, and heterogeneous networks are being deployed where small cells and macrocells now share the same spectrum. With these trends in cellular network evolution, out-of-sector interference is becoming a major impediment. Impairments due to interference are especially severe on the uplink, where near-far effects are more likely to be experienced. This paper provides a holistic view of the network-centric cooperation schemes that have emerged as strong candidates for uplink interference management in evolving cellular networks (e.g., networks based on 3GPP Long Term Evolution based air-interface technologies and beyond). In particular, we introduce three novel approaches: 1) network multiple input multiple output (MIMO), which carries out joint multi-antenna signal processing across sectors; 2) network interference cancellation engine (NICE) which opportunistically cancels dominant interferers that have already been decoded at neighboring sectors and decreases backhaul overhead by one to two orders of magnitude; and 3) a hybrid approach which combines the strengths of both network MIMO and NICE in an attempt to achieve further spectral efficiency without incurring huge backhaul overhead. Several considerations of both theoretical and practical significance (overhead, latency) related to these approaches are discussed, and simpler variants that may apply in the context of heterogeneous networks are also considered. Quantitative investigations of these network-centric cooperation schemes in realistic operating environments show that substantial improvement may be achieved in average and cell-edge spectral efficiency. © 2013 Alcatel-Lucent.