3GPP has introduced support for Internet of Things (IoT) via Cellular IoT (EC-GSM, LTE-M and NB-IoT) standards in its Rel 13 specifications. The aim is to efficiently support billions of devices which vary widely in terms of their processing power, battery life, cost and need for extended coverage. Cellular IoT will support numerous services including utility meters/smart meters, vending machines, automotive (fleet management, traffic management etc), security monitoring and reporting, medical monitoring and alarms/alerts. Depending on services provided, some of these devices will send/receive data infrequently at large intervals (e.g. various sensors/meters in the field) or frequently at regular intervals (e.g. cameras or security appliances).
Cellular IoT supports multiple methods of data transfer between a CIoT-UE and CIoT-Application Server/Services which include the conventional method of data transfer via user plane and new methods such as infrequent and small data transfer via control plane or non-IP data transfer (NIDD) for the efficient use of control plane messaging.
The standards TS 23.401 v13.7.0 and TS 24.301 v13.6.0 (sections 6.2A, 9.9.4.22) define Control Plane CIoT EPS optimization, User Plane CIoT optimization and Non-IP Data Delivery. These methods are appropriate for the transmission of infrequent and small data packets.
For the Control Plane CIoT EPS optimization, data exchange between UE and eNB is done on RRC level and between UE and MME is done at NAS level. 3GPP has recommended ROHC for efficient use of radio resources since Release 4 onwards for WCDMA and since Rel 8 for LTE, mainly for VoIP and other applications on the user plane via PDCP. Now in Rel 13, 3GPP has recommended ROHC for Cellular IoT for NB-IoT services via control plane optimization. The UE and the MME support robust header compression (ROHC) framework if control plane CIoT EPS optimization is supported for PDN connections of IP PDN type.
For uplink IP data, UE implements ROHC compressor, and MME implements the ROHC decompressor. For downlink IP data, MME implements the ROHC compressor, and UE implements the ROHC decompressor. The uplink and downlink ROHC channels are bound by UE and MME to support ROHC feedback. The configurations for the header compression are established during the PDN connection establishment procedure.
The Effnet ROHC product portfolio consists of software products based on various IETF standards which define RObust Header Compression (ROHC).
Effnet ROHC portfolio is integrated in a terminal at NAS level. This proven implementation that has already been integrated into terminals via chipset and protocol stack vendors is highly system efficient, making it possible to run on devices with limited capacity and battery.
Effnet ROHC portfolio is integrated in SGSN-MME at NAS level. This proven implementation that has already been integrated into products, like macro eNB, C-RAN eNB (vBBU), PDSN (from CDMA2000), ASN-GW (WiMAX), is highly system efficient making it possible to scale support for millions of connections/UEs.
Effnet ROHC portfolio is integrated in UE-eNB simulator which supports S1-MME interface testing. This proven implementation that has already been integrated into products, like multi-UE test systems, is highly system efficient making it possible to scale support for millions of simulated connections/UEs.
Effnet’s ROHC product portfolio consists of the following products:
Product | Usage | Profile Identifiers | References |
---|---|---|---|
Effnet ROHC™ | Uncompressed, RTP/UDP/IP, UDP/IP, ESP/IP | 0x0000, 0x0001, 0x0002, 0x0003 | RFC 3095, RFC 4815 |
Effnet ROHC-IP™ | IP | 0x0004 | RFC 3843, RFC 4815 |
Effnet ROHC-TCP™ | TCP/IP | 0x0006 | RFC 4996 |
Effnet ROHCv2™ | Uncompressed, RTP/UDP/IP, UDP/IP, ESP/IP, IP | 0x0101, 0x0102, 0x0103, 0x0104 | RFC 5225 |
Effnet’s ROHC product portfolio has been ported to and integrated on many different platforms. The operating systems include VxWorks, Nucleus, Linux, Windows (2000/XP), Solaris, FreeBSD and processors include PowerPC, MIPS, ARM, SPARC and x86. As the products are highly portable, they can be easily ported to many other operating systems, both real-time and generic as well as to other processors, both 32-bit and 64-bit regardless of byte-order.
In addition to the features specified in the standards, the Effnet ROHC product portfolio has the following efficiency and robustness improving features:
All additional features above are transparent with regard to interoperability.
Packet classification and context management is essential to header compression. Effnet provides this additional module together with the Effnet ROHC product family.
The VoIP enabler on wireless networks! Effnet ROHC™ is an important component to run VoIP services efficiently over wireless networks. Most of the RTP applications use UDP for signaling purposes and there are many stand-alone UDP applications e.g. light weight REST protocols such as CoAP, so the support for IP/UDP compression adds further to the efficiency. There is significant demand for secure exchange of information which leads to increased header overhead. The capability to compress IP/ESP, the header overhead in secure connections, makes it possible to run secure networks without additional bandwidth.
The following table shows some examples of bandwidth savings using ROHC for various codecs and protocols:
Traffic type | Avg. packet size (w/o ROHC) (bytes) | Avg. packet size (w ROHC) (bytes) | Savings (%) |
---|---|---|---|
AMR (4.75Kbps) | 52 | 15 | 71 |
AMR (12.2Kbps) | 71 | 34 | 52 |
IPv4/UDP/CoAP GET/PUT | 53 | 29 | 45 |
IPv6/UDP/CoAP GET/PUT | 73 | 49 | 33 |
As more and more networks are moving to support IP based communications, the number of nodes that require an IP address are increasing rapidly. The introduction of IPv6 should address this concern but at least during transition time, a lot of traffic will be sent via tunnels across networks. Effnet ROHC-IP™ is capable of compressing layers of IP headers thus making it possible to run tunneled traffic without need for additional bandwidth.
Multiple Internet packet size studies* are in agreement that at least 40% of all IPv4 packets carry no or only a few bytes of payload i.e. packet sizes are at or very near to header size (IPv4+TCP). One study of IPv6 packets shows the same trend. Even more remarkable in that study is that for IPv6, 60-80% packets carry more header data than packet data. Effnet ROHC-TCP™ would be very beneficial in these cases.
Traffic type | Avg. packet size (w/o ROHC) (bytes) | Avg. packet size (w ROHC) (bytes) | Savings (%) |
---|---|---|---|
IPv4/TCP ACK (w/o options) | 40 | 10 | 75 |
IPv4/TCP ACK (Timestamp) | 52 | 13 | 75 |
IPv4/TCP HTTP GET (basic) | 70 | 40 | 43 |
* Packet size studies at www.caida.org
Mobility is the cornerstone of the cellular networks but supporting it efficiently is a tricky business. As the cellular network architectures have evolved, the integration point of ROHC in system nodes has moved closer towards mobile terminals for various reasons but has lead to a problem of handling reordering of packets during mobility. Effnet ROHCv2™ addresses this concern very efficiently while providing high compression efficiency and robustness.
The Effnet ROHC product portfolio is offered with a full range of support services, including problem reporting, bug fixes, updates, training, consulting and integration services. A team of engineers experienced in standardization of header compression technology, implementation and testing of product portfolio is available for support and consulting services.
For licensing of the Effnet ROHC product portfolio, complete or individual products, please contact us at info@effnet.com.