Abstract
Future communication networks will be intelligence-embedded on reconfigurable architectures, thus effectively realizing automation of network self-optimization. Central to this architectural approach is handling a comprehensive set of key performance indicators associated with service-specific requirements, such as resource efficiency and operation cost. For heterogeneous networks, node cooperation, which brings computation-intensive intelligence toward cell edges, is important. This article presents a node-cooperative universal framework with a unified class of linear assignment optimizations facilitating massive connectivity and cost-saving user association. Dual self-optimizing targets are classified into two-phase tasks of inter-cell and intra-cell levels. Network nodes share common internal computing structures that can be incorporated in both tasks, and simple message-exchange mechanisms integrate them to achieve orchestration dedicated to specific target configurations. Numerical evaluations analyze impacts of cooperation-based computations, demonstrating cost reduction and efficiency over realistic environments. Within this operational context, we discuss potentials and challenges of network management along with clues to their plausible solutions.
| Original language | British English |
|---|---|
| Pages (from-to) | 50-57 |
| Number of pages | 8 |
| Journal | IEEE Communications Magazine |
| Volume | 62 |
| Issue number | 6 |
| DOIs | |
| State | Published - 1 Jun 2024 |