Mobile operators will need reliable LTE networks to be able to respond to competitive threats, writes Spirent ceo Bill Burns
Mobile operators are sitting pretty, but they are not complacent. Existing 3G networks can match wireline telephony, but they cannot deliver the data and video performance and quality now available on the wireline service via DSL. Moreover, the mobile operators face a challenge from developing wireless technologies – notably WiMAX – that potentially can deliver DSL-type bandwidth or better over wireless networks for mobile subscribers.
The alternative put forward by the mobile operators' 3rd Generation Partnership Project (3GPP) is called 3GPP Long Term Evolution – LTE for short. The targets are for bandwidth as high as 100 Mbps on the downlink, and up to 50 Mbps on the uplink, but this is just a small part of the expected benefits, as LTE is optimised for data traffic and does not carry the burden of a separate, circuit-switched voice network, as in 2G GSM and 3G UMTS networks.
Building on the established 3G system, LTE will begin with the existing transmission towers and backhaul infrastructure in place.
For radio transmission LTE relies on Orthogonal Frequency Division Multiple Access (OFDMA), a technique that combines a division of the signal into time slots with a widely deployed modulation technology called Orthogonal Frequency Division Multiplexing (OFDM). OFDM involves replacing a single high speed stream with multiple slower ones to reduce the effects of signal variation due to Doppler shifts and echo as the receiver moves through the environment. If, instead of transmitting, say, a million bits of data per second on one channel, you transmit a thousand bits per second across a thousand parallel channels, that slower delivery allows better isolation and recognition of each received bit and so a clear data stream emerges even if there is radio distortion in some or all of the channels.
Another key component of LTE development is Multiple Input/Multiple Output (MIMO). MIMO employs more than one antenna to transmit and receive. At the transmission end this is relatively simple, but it requires sophisticated materials to pack multiple antennas into a small receiver. MIMO technology can both be used as a further multiplexing option – with different data streams transmitted simultaneously from different transmit antennas – or, transmitting the same data stream via multiple antennas to improve signal stability. Another more advanced possibility MIMO could provide is, given feedback from a receiver, to use MIMO for real-time shaping of the transmission beam or 'beam forming'.
A significant factor in LTE's success lies in the resulting convergence of technology and networks.More of the LTE system's intelligence is located at the edge of the network – again reducing latency as policy enforcement and decisions are made at the edge rather than referring back to the center.
A typical LTE network relies on two main network elements:
– The base station is an enhanced version called an eNodeB (Evolved NodeB) providing the LTE interface from backhaul to the radio network and providing radio resource management for the evolved access system.
– A new Access GateWay (AGW) is also needed. The AGW provides termination of the LTE bearer signal and acts as a mobility anchor point for the user plane. The AGW is responsible for key logical functions including MME (Mobility Management Entity) for the Control Plane and SAE PDN GW (System Architecture Evolution Packet Data Network GateWay) for the User Plane.
Comparing this with the current 3G architecture, the radio network elements of 3G – such as the Radio Network Controller (RNC) – are now shared between the AGW and the eNodeB, while the core network functions – such as SGSN and GGSN or PDSN (Packet Data Serving Node) and rthe outers – are distributed mostly towards the AGW.
LTE Security
Another very important part of the 3GPP's strategy is to embed security into the LTE network using a multi-layer, multi-vendor approach rather than fall back on patching holes in an ad hoc manner.
Moving to an all-IP network is one of the most significant security challenges.
An end-to-end system approach to security is being developed. It begins with initial hardening of the LTE platform to make sure it is without loopholes and to secure the integrity of the software and its configuration. Then the system must include user and operator authentication, authorization and auditing systems. As well as a secure protocol, network management and data storage need to be secured. All these security issues must apply both to protect the perimeter of the network and to seek out and forbid unsolicited traffic within the network.
Business requirements for LTE testing
Another key requirement that needs to be tested and monitored is the quality of experience – users that were persuaded to upgrade from 2G to 3G won't want to upgrade to a frustrating new service.
From the operator's point of view, a service oriented architecture (SOA) is essential for future-proofing their LTE investment. Users will expect a flow of new services to be provided to their handsets without delay or hardware upgrades. SOA opens the network to rapid roll-out of third party applications as fast as they are created – and that builds customer loyalty.
Content-based charging is another very desirable feature. To be able to adjust pricing so that, say, premium high definition video services can be charged at a premium, and to have the flexibility to adjust individual service charges to meet competition in the market.
Testing LTE infrastructure
The criteria for LTE testing are not that different from those required for 2G services but, as a new technology, LTE demands more stringent testing to iron out any unexpected flaws and build market confidence. Pre-deployment testing ensures that the operator can launch a fully tested service that meets the criteria of the previous section and offers optimal quality of experience (QoE) for subscribers from day one. As subscriber numbers increase and greater pressure is put on the system, the same tests need to be maintained to ensure that performance and QoE standards are maintained for the growing user base, and to monitor for any unanticipated sources of service degradation. Pre-testing services reduces time-to-market for new services, and post testing reduces subscriber churn
As well as testing broadband performance, throughput, reliability, recovery, packet loss and other factors, the operator needs to anticipate problem areas and test all these factors under less than ideal conditions such as network overload transmission to and from a high speed train, performance in a cluttered urban environment, signal deterioration under poor RF coverage and so on.
Testing devices for the LTE air interface
Test devices must emulate real-world radio channel challenges – such as time-varying multipath delay spread, fast fading, slow shadow fading, radio noise and channel loss – so that developers can put devices through their paces at the test stage and identify performance issues early in the development and design verification cycle. Artificial challenge conditions can be invented, or the testers can re-create in the lab extreme conditions that were actually met and measured in the field.
Replicating a realistic RF environment for testing has significant advantages over relying on an actual over-the-air signal. For one thing, the environment is repeatable, so test iterations provide "apples to apples" comparisons. For another, an emulated environment can be easily and precisely controlled, so the test platform encourages "what-if" testing in anticipation of possible future challenges:
– What happens as a parameter changes?
– What will happen if we roll out this new data-centric service?
– What happens as multiple parameters or impairments change?
– What happens as a device moves from cell to cell
– What happens as it moves between 2G and 3G networks
Each new generation of wireless further stresses radio-link performance, and LTE is no exception. Higher data rates mean higher sensitivity to environmental radio effects, and while MIMO and beam forming may not be deployed in the first LTE rollouts, the wireless links used by these advanced antenna techniques are extremely complex.