Description
The Buffer Status Report (BSR) is a Medium Access Control (MAC) protocol data unit (PDU) control element defined in 3GPP specifications for LTE (E-UTRA) and NR (New Radio). Its primary function is to provide the network (specifically, the gNB in 5G NR or eNB in 4G LTE) with information about the amount of uplink data queued for transmission in the UE's buffers, categorized by logical channel groups (LCGs). This information is essential for the network's uplink scheduler to make intelligent decisions about resource allocation (Resource Blocks, RBs). The BSR mechanism operates within the MAC sublayer of the UE's protocol stack. When data arrives at the UE's Packet Data Convergence Protocol (PDCCP) layer for a logical channel, it is stored in corresponding buffers. The MAC layer monitors these buffers and triggers the generation of a BSR under specific conditions defined by standardized triggering events.
Architecturally, the BSR is transmitted as a MAC Control Element (CE), which is multiplexed with MAC Service Data Units (SDUs) from higher layers to form a MAC PDU for transmission on the Uplink Shared Channel (UL-SCH). There are several types of BSRs defined to optimize signaling overhead and responsiveness: Regular BSR, Periodic BSR, and Padding BSR. A Regular BSR is triggered when uplink data becomes available for a logical channel that has higher priority than those currently with data, or when no data is available for any logical channel and then data arrives (this triggers a BSR for an 'empty buffer'). A Periodic BSR is triggered by the expiry of a timer (periodicBSR-Timer), ensuring the network receives updates even during sustained data flows. A Padding BSR is triggered when the UE has uplink resources allocated, but the amount of padding bits required to fill the transport block would be sufficient to carry a BSR MAC CE, allowing for efficient use of granted resources.
The content of the BSR MAC CE includes a Buffer Size field for each configured Logical Channel Group (LCG). In LTE and NR, up to 8 LCGs can be configured (0-7), though typically 4 are used. The Buffer Size indicates the total amount of data available across all logical channels belonging to that LCG, including data that is available for transmission in the RLC and PDCP layers. The reported value is an index that maps to a range of bytes (e.g., 0 bytes, 1-10 bytes, 11-20 bytes, etc.), as defined in a lookup table in the specifications (TS 36.321, 38.321). This quantization reduces signaling overhead. Upon receiving a BSR, the gNB/eNB's scheduler uses this information, along with other factors like UE channel quality (CQI), QoS requirements of the bearers, and overall cell load, to decide the size, timing, and frequency of uplink grants (via the Uplink Grant in DCI format 0_1 in NR or DCI format 0 in LTE) sent to the UE on the Physical Downlink Control Channel (PDCCH).
The BSR procedure is tightly coupled with other MAC procedures like Logical Channel Prioritization (LCP), which determines which logical channel's data is placed into a MAC PDU once a grant is received. A key timer, `retxBSR-Timer`, ensures reliability; if a BSR is triggered but the UE has no uplink resources to send it, it starts a Scheduling Request (SR) procedure. If the SR is not successful before the timer expires, the BSR is retriggered. This closed-loop feedback between UE buffer status reporting and network scheduling is fundamental to the dynamic and efficient operation of the uplink in both LTE and 5G NR, enabling support for diverse traffic types from low-latency small packets to high-throughput data streams.
Purpose & Motivation
The BSR was introduced to solve the fundamental problem of efficient uplink resource allocation in a shared channel system like LTE and NR. In earlier cellular systems (e.g., 3G WCDMA), uplink transmissions were code-scheduled or contention-based, which could lead to inefficiencies under dynamic traffic loads. The shift to OFDMA/SC-FDMA in LTE required a more sophisticated, grant-based uplink where the network must explicitly tell the UE when and on what resources to transmit. For this to be efficient, the network scheduler needs accurate, timely knowledge of each UE's transmission needs. Without a BSR-like mechanism, the network would have to blindly allocate resources, either wasting them (over-granting) or causing excessive queuing delays (under-granting), severely impacting QoS, latency, and overall system capacity.
The creation of the BSR mechanism was motivated by the need to support diverse QoS requirements and packet sizes inherent in IP-based services. Applications generate bursty, unpredictable traffic. A static or purely periodic grant allocation cannot adapt to this variability. The BSR provides the essential feedback for dynamic scheduling, allowing the network to allocate resources 'on-demand.' This is particularly critical for services with strict latency bounds (e.g., VoIP, gaming) where data must be transmitted immediately upon arrival, and for best-effort services where throughput should be maximized when data is available. The BSR, therefore, is a cornerstone for enabling the packet-switched, all-IP architecture envisioned for 4G and 5G, moving away from the circuit-switched mindset of previous generations.
Furthermore, the BSR design addresses the trade-off between reporting accuracy and signaling overhead. By grouping logical channels into LCGs and using quantized buffer size levels, it minimizes the bit-length of the report. The different trigger types (Regular, Periodic, Padding) optimize for different scenarios: immediate need, periodic updates, and opportunistic reporting. This elegant design ensures that the uplink scheduling loop is both responsive and efficient, a key factor in achieving the high spectral efficiency and low latency targets of LTE and NR systems.
Key Features
- Reports uplink buffer occupancy per Logical Channel Group (LCG) to the gNB/eNB scheduler
- Transmitted as a MAC Control Element (CE) on the UL-SCH
- Defines multiple trigger types: Regular, Periodic, and Padding BSR for optimal signaling
- Uses quantized buffer size indexes to minimize overhead
- Integrates with Scheduling Request (SR) procedure for resource acquisition when needed
- Essential for dynamic uplink scheduling enabling QoS-aware resource allocation
Evolution Across Releases
Introduced the fundamental BSR framework for LTE (E-UTRA). Defined the BSR MAC Control Element formats (short and truncated), the three triggering events (Regular, Periodic, Padding), and the association of logical channels to up to 4 Logical Channel Groups (LCGs). Established the core procedures involving timers (periodicBSR-Timer, retxBSR-Timer) and interaction with the Scheduling Request mechanism.
Enhanced BSR for Carrier Aggregation (CA). Introduced the need for BSR reporting per cell group in scenarios with multiple serving cells. The MAC layer was extended to handle buffer status reporting across aggregated carriers, ensuring the scheduler has a unified view of UE buffer status for optimal multi-carrier scheduling.
Introduced Dual Connectivity (DC). This required enhancements to BSR procedures to support buffer status reporting for two separate cell groups (Master Cell Group and Secondary Cell Group) managed by different nodes (MeNB and SeNB). Separate BSR processes and reporting were defined for each cell group.
Defined the BSR framework for 5G New Radio (NR). Maintained the core principles from LTE but adapted for the NR MAC layer. Key updates included support for up to 8 Logical Channel Groups (LCG ID 0-7), new MAC PDU formats, and alignment with the more flexible NR frame structure and bandwidth parts. BSR remains critical for dynamic grant and configured grant (semi-persistent scheduling) uplink schemes.
Enhanced BSR for NR-based operation in unlicensed spectrum (NR-U). Introduced considerations for listen-before-talk (LBT) and channel access procedures, which could impact the timing and reliability of BSR transmission. Optimizations were studied to ensure efficient scheduling despite the uncertainties of unlicensed channel access.
Further enhancements for NR, including support for Integrated Access and Backhaul (IAB). For IAB nodes, which act as both UE and gNB, BSR procedures were extended to handle the distinct buffer status for access and backhaul links, ensuring proper scheduling across the multi-hop network topology.
Defining Specifications
| Specification | Title |
|---|---|
| TS 21.905 | 3GPP TS 21.905 |
| TS 36.300 | 3GPP TR 36.300 |
| TS 36.321 | 3GPP TR 36.321 |
| TS 36.842 | 3GPP TR 36.842 |
| TS 38.321 | 3GPP TR 38.321 |
| TS 38.331 | 3GPP TR 38.331 |
| TS 38.401 | 3GPP TR 38.401 |