Explain how network slicing is implemented in 5G architecture.

Network slicing creates multiple virtual end-to-end networks on shared physical infrastructure. Implementation layers: RAN slicing: Allocates radio resources per slice through scheduling and resource reservation. Transport slicing: Uses SDN/NFV for traffic isolation, QoS differentiation. Core slicing: Dedicated or shared network functions per slice; AMF, SMF, UPF can be slice-specific. Key identifiers: S-NSSAI (Single Network Slice Selection Assistance Information) identifies slices; contains SST (Slice/Service Type) and optional SD (Slice Differentiator). Selection process: UE requests slice(s), AMF validates and selects based on subscription and availability. Slice types: eMBB (high throughput), URLLC (low latency), mMTC (massive IoT). Challenges: Resource isolation guarantees, slice lifecycle management, cross-slice optimization, and SLA enforcement. Orchestration uses NFV MANO frameworks.