Mobile Backhaul: The Missing Link

James P. Cavanagh

Mobile is hot and mobile backhaul is thriving. As the number of mobile devices grows the need to bridge the gap between the subscriber devices and the IP cloud grows and as the bandwidth demand increases the demand for a bigger bridge between the subscriber devices and the IP cloud also grows. This is where “mobile backhaul”, or as it is sometimes called “wireless backhaul”, comes into play. Backhaul is independent of the endless bickering over what the “G” du jour might be, “2.5G”, “3G”, “4G”, or whatever. And, the backhaul technology choice is independent of the air interface because when the bits are handed off to the backhaul to be transported from the tower to the Internet they are rendered as packets: the same format as every other packet anywhere in the Internet.

“Deja vu All Over Again!”

I can’t help but get the feeling I’ve been here before, but with a twist. I understand the way many of the American colonists felt: they lived in a brave new world, a world full of promise and a better way of doing things, but they were still loyal to their homeland and reluctant to cast aside all of their British culture and fully embrace the new ways being given life all around them. I, too, am a colonist: packet colonist in a brave new world who has not completely left the world of circuits and Time Division Multiplexing. The last time I felt as I do now the dust was settling after the “Circuit vs Packet” wars. Packet had clearly won, Vint Cerf and his colleagues were taking their victory laps at conferences and trade shows and it was a “brave new world” in global networking! The migration to packet which many organizations, such as DoD, had begun a quarter of a century before was now a stampede and everyone wanted one of Dr. Cerf’s famous “IP on Everything” tee shirts.

What most people, except for the Layer 1 engineers, failed to grasp is that circuits – which had been defeated so severely by packets – were still there! It is impossible for packets to move without the bit path, timing and synchronization provided by the underlying physical circuits. And, those circuits, to a very large extent even today, rely on the strict formatting, transmission and timing rules of SONET or SDH, which are high-speed versions of TDM and which is still the realm of the old circuit regime. So, why do I feel I have been here before? Because in the early 1990s packets “won” but still required underlying “circuits” and today mobile with packets, has “won” but still requires underlying circuits. What’s the twist? Mobile packets go from air interfaces which transport packets; the packets must then travel across the backhaul where there are often serious bandwidth constraints, and often use SONET or SDH rules to carve the bandwidth into circuits, and then are let loose into the packet realm once again where bandwidth is cheap and plentiful. This is a case of the least common denominator. The least good handling is the best that the connection will get.

The Bottleneck

With every passing day more and more people, and their apps, are untethered from wires and are becoming a part of mobile mania. The transition requires more and more bandwidth in more and more places and, at least until the big wireless companies get caught up, that gap is being filled by a wide variety of contractors and sub-contractors with a wide variety of technologies, reliabilities, Service Level Agreements, Qualities of Service and Qualities of Experience. The challenge for the big carriers is to assemble all of the piece parts at a cost which allows them to make money while prices fall and provide a good, consistent service quality nationwide while avoiding customer attrition based on poor service quality. And, that bottleneck is consistently high-capacity, high availability mobile backhaul.

Why is this so problematic? Let’s not even address the variations between markets that can exist in a single nationwide carrier. Most nationwide carriers have lacked the luxury of building out their networks from the ground up. Their networks are often stitched together through the acquisition of dozens of smaller companies each with their own disparate technologies, vendors and architectures. So we will leave that discussion for another time and, instead, let’s focus on the variability that is possible in wireless backhaul, the thing I am referring to as “the missing link”. The performance of packets on wireless networks is pretty well understood by wireless engineers and their performance works in a manner consistent with wireless standards. Very few carriers, in fact, have multiple wireless technologies in their networks and those that do are in the process of migrating to a single system-wide technology, so there is consistency in that part of the end-to-end connection. Once a packet hops off of the backhaul it is usually only one or two hops from the Internet, often on a private IP-based intranet and that part of the packet’s trip follows IP rules, also reasonably well understood. But, the middle piece, the backhaul, very often employs circuit technology which is formatted to the rules of SONET and SDH. This packet-circuit-packet translation is often the problem but also has less well understood benefits. Why?

Think about the highway system for just a moment. Imagine a big interstate highway where any traffic can travel in any one of eight lanes, dynamically, as traffic requires. That would represent all of the wireless calls being delivered dynamically on a reasonably well-engineered wireless network. Just like the interstate example the wireless network can fill up, slow down calls in progress and not allow new calls to start but, just like the interstate, the wireless network is engineered to provide very good service most of the time. Now, picture a bridge installed on the eight lane interstate that has six lanes, all of which are designed to allow a certain amount of traffic at a fixed speed with all of the vehicles kept in order. How does this impact critical characteristics of the traffic, such as delay, delay variation, discard and availability? Well, if the sum of the “bandwidth” of the lanes on the bridge exceeds the “bandwidth” of the interstate and the rules for who gets to use which lane are set up correctly, the bridge could actually improve traffic characteristics, especially if the vehicles were accepted onto the bridge and traveled across the bridge at a rate that compensated for the 8:6 ratio of highway lanes to bridge lanes. If the bandwidth of the bridge were less than the bandwidth of the interstate the presence of the bridge would hurt traffic characteristics when there was a lot of traffic. But, another element comes into play. Because of the structured nature of the bridge the traffic flow and performance actually become more predictable (a characteristic associated with quality) and an opportunity is created to set up a prioritization system with higher value traffic being given preference. In the case of the bridge this would involve toll booths, high priority lanes and other such traffic management mechanisms. In the case of the network example this would involve Type of Service bits, Virtual LAN IDs and similar mechanisms coupled with Service Level Agreements and specialized enforcement and billing arrangements. These benefits are often overlooked because the “system” is being developed and implemented by packet people who lack the understanding of the benefits of circuit.

In other cases the “system” is being developed and implemented by circuit people who don’t appreciate or understand packets. The packet environment really is different, with concepts that are foreign to circuit people, such as the idea that a structured series of bits, a packet, could be intentionally discarded by the network because a specific user has given the network too many packets to deliver within a certain timeframe. In this case packet benefits are overlooked.

Conclusion

At this moment most organizations are trying to go from packet to packet over a bridge built with circuit technology. This can be problematic or rich with advantages, as outlined earlier. What is the future? Any savvy organization will forget about the technology wars and will seek to employ, or continue to employ, the best parts of circuit and packet technology to further their own business objectives. Depending upon those objectives the ultimate solution might be all packet or a packet-circuit hybrid. I think the days of an all-circuit solution are both years behind and years ahead, but that is a discussion for another time.

As for myself, I don’t think it matters if I am a circuit guy living in a packet world or a packet guy who was born in the circuit world. What matters is that I know both worlds intimately. I can be the perfect ambassador for either one or can broker a lasting peace between the two worlds that leverages the unique characteristics of each for the benefit of all. Could I actually be the missing link? It makes one stop and think, doesn’t it?

Editor’s Note: Besides 30+ years of experience that includes hands-on engineering, consulting, and training on a wide range of networking technologies, Jim Cavanagh, a Principal Member of our Telecommunications Faculty, is our resident expert on mobile backhaul as well. He has taught numerous classes for clients using a wide range of backhaul options — including traditional SONET/SDH, multiple broadband technologies, satellite, and other wireless technologies — to get the multimedia traffic from their users’/clients’ hands to the ‘net. He has instructed backhaul engineers in the proper set-up of multiple backhaul devices; advised large multinationals, government, and public safety agencies on backhaul security and the use of backhaul as a part of VPN and Cloud-based networks; documented backhaul best-practices; helped establish SLAs and service descriptions; and even written backhaul marketing materials. He shares his extensive, practical knowledge of wireless backhaul with our clients via onsite and WebLive™ classes that are part of the Eogogics Mobile Wireless Backhaul Curriculum. See Jim’s Detailed Bio.