Jan 7, 2015
The commonly used means of providing internet access to remote or small offices is via a wired (or fiber) connection. From ISDN through ADSL to T1 and DS3 lines, much money has been spent on provisioning and operating such lines. However, times are changing, and rather than regarding an organization as a grouping of central, remote and home “offices”, it is not necessary to regard the organization as a collection of individuals who could be working from anywhere at any time – whether this is from a BYOD device on the road or from some office location somewhere.
The use of high-cost physical links is becoming a constraint on how a business can operate. Wireless has long been accepted as being a flexible solution – but high cost and low bandwidth have generally been seen as being show-stoppers in planned usage.
However, the times they are a changing, as Dylan once said. Long-Term Evolution (LTE) is a catch-all term for the current generation of wireless wide area 4G network technologies.
The original 4G specifications were for LTE to support up to 300Mb/s download speeds and 75Mb/s upload; however, LTE Advanced allows for systems capable of reaching 1Gb/s and possibly greater speeds. Through a simple agreed redefinition of the original LTE requirement, LTE and LTE Advanced (along with other new generation packet switch wireless WAN approaches) have all been clumped together under a generic term of “4G LTE”. For the purposes of this article, LTE will apply to the orthogonal frequency division multiple access (OFDMA) system as used by the majority of new mobile devices today – it will not include WiMAX or HSPA+. It is also argued by purists that existing LTE technology is not suitable for “pure” 4G and that LTE Advanced is required in order to meet the letter of the ITU-R 4G definition. In real use, LTE is good enough to meet most needs. LTE advanced, however, is an evolution of LTE – in the majority of cases, what applies to LTE will also apply to LTE Advanced. Although it sounds as if the current state of LTE is a mess, it is not as bad as it as first appears.
4G LTE was first implemented in a commercial mobile device in 2009, and is likely to be around at least until the 2020s, as the next generation of mobile wireless connectivity (5G) has not been agreed to any research depth yet. 4G LTE was 5 years in research before it hit the markets: even if a solid 5G research program were put in place today, it would be likely to be 2020 before 5G reached the market, and even then it is likely to be more backwards compatible than previous “G”s have been.
What this means for service providers and users is that 4G LTE is here, it is proven, and it is likely to be around for long enough to give good returns on investment. With decent up and down bandwidth, 4G LTE can be used for the vast majority of data needs: it can work with real-time bi-directional data needs such as video and voice (although a move from circuit switching to packet switching technologies has created head-scratching for some providers until a workable and usable Voice over LTE (VoLTE) approach was agreed as a general direction) as well as high-speed data transfers such as large files, particularly in video streaming. Where ultra-high bandwidth is required, bonding of wireless WAN systems can be used – as long as the base station and backhaul are up to it. It is also optimized for true mobility – handover between 4G LTE base stations and Wi-Fi access points can be carried out for devices traveling at 220mph or 310mph, depending on the frequency band being used.
As a wireless technology, organizations can gain access to LTE networks far faster than waiting for cables to be connected to their facilities: time to capability is massively improved and security of supply is enhanced, as the wiring via a single point is no longer an issue. For example, according to the Open Networking User Group (ONUG), provisioning a T1 MPLS line can take between 30 and 90 days, with fibre lines taking even longer. 4G LTE WAN is generally “there” – as a wireless technology, all that is required is suitable customer premise equipment (CPE) to create the connection.
However, the main benefit in deploying LTE as a platform is that it provides the capability for the platform to be more easily software defined. LTE is packet-based, moving away from the old problem of circuit-switched WAN networks. Data can be regarded as being essentially IP-based and dealt with in a more cohesive end-to-end manner, and software used to monitor, manage and take actions on the data.
Data being sent across a 4G LTE network can be monitored and managed across the whole transport in a far more transparent manner than currently where a mix of packet and circuit switch systems are being utilised with completely different standards in place.
Sure – current 4G LTE availability is patchy, but service providers will follow the money. Once the demand is there, they will provide: for organizations requiring a more flexible means of connecting remote workers and central offices via a public WAN, 4G LTE managed through SD-WAN tooling will provide far better economics and capabilities than physical connections.
This post is part of an ongoing series examining the issues facing enterprises seeking to implement a Software-Defined WAN (SD-WAN) solution, as addressed in the Open Networking User Group white paper, “ONUG Software-Defined WAN Use Case”.