Gateway Packet Splitting for LTE HetNet Latency Reduction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current LTE HetNet deployments face increased data plane latency due to packet forwarding over non-ideal backhaul interfaces between master and secondary access nodes, limiting bandwidth and degrading user device performance.

Innovation Solution

A gateway that splits data packets into sub-sequences and forwards them directly to both master and secondary access nodes, reducing latency and enhancing bandwidth by avoiding excessive packet forwarding through the master node.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If packet forwarding is performed over the X2 interface between master and secondary access nodes, then control plane connectivity is maintained, but data plane latency increases due to non-ideal backhaul

Engineering Contradiction:
Improvecontrol plane connectivityVSAvoiddata plane latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the data plane and control plane connections. The control plane connection remains anchored to the master access node via X2 interface, while the data plane establishes direct connections from the gateway to both master and secondary access nodes. This segmentation allows control signaling to flow through the master node while user data packets are transmitted directly to secondary nodes, eliminating the latency penalty of routing all traffic through the X2 interface.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The gateway acts as an intermediary that receives data packets from the core network and performs packet-level RAB splitting, directly forwarding packets to both master and secondary access nodes without requiring intermediate forwarding through the X2 interface. This intermediary function eliminates the need for secondary nodes to queue packets received from the master node, reducing data plane latency.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If RAB splitting is performed at RAB granularity by the S-GW (1A architecture), then QoS requirements are met, but aggregated bandwidth is limited since only a single eNB schedules transmissions

Engineering Contradiction:
ImproveQoS requirementsVSAvoidaggregated bandwidth
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies segmentation at the packet level within each RAB, dividing data packets into multiple sub-sequences that can be concurrently transmitted through multiple access nodes. This allows a single RAB to utilize the aggregated bandwidth of both master and secondary eNBs while maintaining QoS requirements through controlled packet distribution and reassembly at the user device.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from single-dimension scheduling (one eNB per RAB) to multi-dimensional scheduling by enabling concurrent transmissions through multiple access nodes for the same RAB. This dimensional expansion allows the system to simultaneously satisfy QoS requirements through controlled packet routing while maximizing aggregated bandwidth through parallel transmission paths.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Productivity

If packet-level RAB splitting is performed at the MeNB (3C architecture), then bandwidth aggregation is improved, but data plane latency increases due to packet forwarding over X2 interface

Engineering Contradiction:
Improvebandwidth aggregationVSAvoiddata plane latency
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent extracts the packet forwarding function from the master access node by implementing packet-level RAB splitting at the gateway instead. The gateway directly forwards packet sub-sequences to the secondary access node, removing the intermediate forwarding step through the X2 interface that causes latency. This extraction eliminates the bottleneck while preserving bandwidth aggregation benefits.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs packet-level RAB splitting in advance at the gateway before packets reach the access nodes. By pre-dividing packets into appropriate sub-sequences and routing them directly to the correct nodes, the system eliminates subsequent forwarding operations that would introduce latency at the master node, thereby reducing overall data plane latency while maintaining bandwidth aggregation.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS10757606B2Gateway, master access node and methods thereof
Publication Date: 2020.08.25 HUAWEI TECH CO LTD
  • US10757606B2 patent drawing
  • US10757606B2 patent drawing
  • US10757606B2 patent drawing

AI summary

A gateway comprises a transceiver configured to receive a sequence of data packets on a Radio Access Bearer, RAB, from a core network, a processor configured to obtain a first distribution of the received sequence of data packets into a first sub-sequence of data packets and at least a second sub-sequence of data packets, wherein the transceiver is configured to forward the first sub-sequence of data packets on the RAB to a master access node, forward the second sub-sequence of data packets on the RAB to a secondary access node. The master access node comprises a processor, a transceiver configured to receive a first sub-sequence of data packets of a sequence of data packets carried by a RAB from a gateway, wherein the RAB is associated with a user device, and to forward at least one part of the first sub-sequence of data packets on the RAB to the user device.