Header Compression — Architecture & Use Cases

Overview

Header compression reduces the size of protocol headers to improve bandwidth efficiency, reduce transmission overhead, and lower latency—especially on links where header size is large relative to payload size. By exploiting redundancy across packets in the same flow, compression mechanisms transmit only the fields that change, while reconstructing the full header at the receiver.

Header compression is essential in wireless networks, low‑bandwidth links, IoT systems, and real‑time applications where every byte matters. Modern schemes balance compression efficiency, robustness, and computational cost.

Why Header Compression Matters

Header compression addresses several challenges:

High overhead — IP/UDP/RTP headers can exceed 40 bytes, which is significant for small payloads.
Low‑bandwidth links — wireless, satellite, and IoT links benefit from reduced overhead.
Latency sensitivity — smaller packets reduce transmission time and jitter.
Energy efficiency — fewer transmitted bits reduce power consumption in mobile devices.
Scalability — compressed headers reduce load on constrained networks.

Compression is especially impactful for real‑time voice/video, sensor data, and tunneling protocols.

Architectural Principles

Field Classification

Headers contain fields with different behaviors:

Static fields — rarely change (IP version, source address).
Semi‑static fields — change infrequently (TTL, flow label).
Dynamic fields — change every packet (sequence numbers, timestamps).

Compression exploits this structure by sending only what is necessary.

Context Maintenance

Both sender and receiver maintain a shared context describing:

static header fields
inferred values
prediction rules
last‑seen dynamic fields

Context synchronization is critical for correctness.

Delta Encoding

Dynamic fields are encoded as:

increments (e.g., +1 sequence number)
differences from previous values
predicted values with correction bits

Delta encoding dramatically reduces header size.

Robustness Mechanisms

Compression must tolerate:

packet loss
reordering
context desynchronization

Schemes include periodic refreshes, negative acknowledgments, and fallback modes.

Compression Techniques

Stateless Compression

No context is maintained; compression is limited to simple transformations.

Pros: robust, simple
Cons: low compression ratio

Used in simple link‑layer protocols.

Stateful Compression

Sender and receiver maintain synchronized context.

Pros: high compression ratio
Cons: sensitive to loss; requires context repair

Used in ROHC, CRTP, and modern wireless systems.

Predictive Compression

Predicts dynamic fields based on flow behavior.

Examples:

RTP timestamp increments
TCP sequence number increments
IPv6 flow label patterns

Prediction reduces the number of transmitted bits.

Dictionary‑Based Compression

Uses shared dictionaries to compress repeated patterns.

Used in:

6LoWPAN
IoT protocols
header compression for constrained networks

Key Header Compression Standards

ROHC (Robust Header Compression)

Designed for wireless links with high loss and variable delay.

Supports compression of:

RTP/UDP/IP
UDP/IP
TCP/IP

Features:

strong robustness
multiple compression modes
context repair mechanisms

Widely used in LTE, 5G, and VoIP over wireless.

CRTP (Compressed RTP)

Compresses RTP/UDP/IP headers from ~40 bytes to 2–4 bytes.

Pros: high efficiency
Cons: sensitive to loss; limited robustness

Used in early VoIP systems.

IPHC (IP Header Compression)

Compresses IPv6 headers for low‑power networks.

Used in 6LoWPAN and IoT systems.

ECRTP (Enhanced CRTP)

Improves robustness over lossy links.

Adds:

context repair
improved loss tolerance
better handling of reordering

Use Cases in Real Systems

Wireless Networks (LTE, 5G)

Header compression reduces overhead for:

VoLTE
real‑time video
IoT traffic
tunneling protocols

ROHC is the dominant mechanism.

IoT and Low‑Power Networks

6LoWPAN compresses IPv6 headers to a few bytes.

Benefits:

reduced energy consumption
improved reliability
support for constrained devices

Real‑Time Voice and Video

RTP header compression reduces latency and jitter.

Used in:

VoIP
WebRTC
conferencing systems

Tunneling and VPNs

Encapsulation adds multiple headers; compression mitigates overhead.

Examples:

GRE
IPsec
VXLAN
GENEVE

Compression improves efficiency in overlay networks.

Performance Considerations

Compression Ratio

Depends on:

header structure
flow predictability
context stability
packet loss rate

Stateful schemes achieve the highest ratios.

Latency

Compression adds processing overhead.
However, reduced packet size lowers transmission delay.

Robustness

Lossy links require:

periodic context refresh
fallback to uncompressed mode
negative acknowledgments

ROHC is designed specifically for robustness.

Energy Efficiency

Smaller packets reduce transmission energy, critical for IoT and mobile devices.

Comparison of Compression Schemes

Scheme	Compression Ratio	Robustness	Complexity	Typical Use
Stateless	Low	Very High	Low	Simple links
CRTP	High	Low	Medium	Early VoIP
ECRTP	High	Medium	Medium	Wireless VoIP
ECRTP	Very High	Very High	High	LTE/5G
6LoWPAN IPHC	Medium	High	Low	IoT

Design Tradeoffs

Compression ratio vs robustness — aggressive compression is fragile under loss.
Stateful vs stateless — stateful improves efficiency but requires context repair.
Processing cost vs energy savings — compression consumes CPU but reduces transmission energy.
Flexibility vs complexity — programmable compression supports new protocols but increases overhead.
Latency vs efficiency — compression reduces transmission time but adds processing delay.

Summary

Header compression reduces protocol overhead by exploiting redundancy across packets, enabling efficient communication over wireless, low‑bandwidth, and latency‑sensitive links. Through stateful context maintenance, delta encoding, and robust repair mechanisms, modern schemes like ROHC and 6LoWPAN achieve high compression ratios while maintaining reliability across diverse environments.

DexterLab