Adaptive Bandwidth Part (BWP)

The core mechanism of BWP allows a UE to switch between up to four pre-configured bandwidth parts on a serving cell. A BWP is defined by a specific numerology (subcarrier spacing) and a contiguous set of Physical Resource Blocks (PRBs). Switching is triggered by Downlink Control Information (DCI) for fast adaptation, a BWP inactivity timer for automatic fallback, or RRC signaling for initial configuration. This allows the UE to operate on a narrow BWP for paging and monitoring, then instantly switch to a wide BWP for a burst of high-throughput data.

A critical power-saving mechanism that automatically transitions the UE from an active, wide-bandwidth BWP to a default, narrow-bandwidth BWP after a configurable period of scheduling inactivity. The timer is reset upon every transmission or reception on the active BWP. Upon expiry, the UE falls back to the default BWP, which is typically configured with minimal bandwidth to conserve power. This ensures that the UE does not remain in a high-power, wide-bandwidth state unnecessarily, directly reducing its energy consumption without explicit signaling overhead.

Each BWP can be configured with a distinct numerology, defined by its subcarrier spacing (e.g., 15 kHz, 30 kHz, 60 kHz). This enables a single carrier to simultaneously support diverse service types. A narrow BWP with a wider subcarrier spacing can be optimized for Ultra-Reliable Low-Latency Communication (URLLC) with short slot durations, while a wide BWP with a narrow subcarrier spacing is optimized for enhanced Mobile Broadband (eMBB) throughput. This is a key enabler for efficient network slicing on the same physical carrier.

When a UE switches its active BWP, it must retune its RF front-end and adjust its baseband processing to the new center frequency and bandwidth. This process requires a small interruption known as a retuning gap. The 3GPP standard specifies the duration of this gap based on UE capability and the frequency range. Sophisticated scheduling algorithms must account for this gap to avoid scheduling data during the transition, ensuring seamless connectivity. Efficient retuning is critical for realizing the energy savings of BWP without degrading user experience.

BWP is a powerful tool for implementing energy-aware slice selection. A network slice configured for massive IoT can be mapped to a narrow, low-power BWP, while a slice for high-definition video streaming is mapped to a wide BWP. The Slice-Aware MAC Scheduler can then dynamically assign users to the most energy-appropriate BWP based on their active slice's SLA. This coordination between BWP and network slicing minimizes the overall power footprint of the RAN by ensuring that each service class uses only the necessary spectral and power resources.

Beyond the inactivity timer, BWP adaptation is a cornerstone of the 3GPP Release 16 UE power saving study. Techniques include: - DCI-based switching: The gNB explicitly commands the UE to a narrow BWP for low-activity periods. - Bandwidth part switching delay: Defined UE capability signaling ensures the network knows the exact time a specific UE model requires to switch, preventing scheduling errors. - Cross-slot scheduling: Combined with BWP switching, this allows the UE to skip PDCCH monitoring on certain slots, entering a micro-sleep state for further power reduction.

A BWP is defined by its numerology (subcarrier spacing), location (starting PRB), and bandwidth (number of PRBs). A UE can be configured with up to 4 BWPs per carrier, but only one is active at a time. The network signals BWP switches via DCI (scheduling-based) or RRC (timer-based) reconfiguration, allowing seamless transitions between a wide BWP for throughput and a narrow BWP for power saving.

A critical power-saving mechanism that automatically switches the UE from an active wide BWP to a default narrow BWP after a configurable period of data inactivity. This timer is reset with every data transmission or reception. When the timer expires, the UE falls back to the narrow BWP, reducing its RF bandwidth and baseband processing, which directly lowers power consumption without explicit scheduling commands.

Cell DTX is the network-side counterpart to UE BWP. When a cell has no active UEs, it periodically suspends transmission of always-on signals like SSB and SIB1. Combined with BWP, this creates a system-wide power-saving state: UEs operate on narrow BWPs with minimal monitoring, while the gNB enters low-energy transmission modes. This synergy is essential for meeting Net Zero targets in dense urban deployments.

A WUS is a low-complexity, narrowband signal transmitted by the gNB to alert a UE in deep sleep that it must wake up to monitor the PDCCH. When combined with BWP, a UE can remain on a minimal BWP and only need to detect a simple WUS instead of performing full PDCCH blind decoding. This drastically reduces the active time and processing load for IoT and mMTC devices.

While BWP reduces the UE's active bandwidth, Resource Block Muting reduces the gNB's transmission bandwidth by deactivating power on unscheduled PRBs. A gNB serving multiple UEs on narrow BWPs can mute the remaining PRBs in the carrier, achieving dynamic bandwidth adaptation on the network side. This technique is a key enabler for energy-proportional RAN where power scales linearly with traffic load.