Will the Switch to IPv6 Create VoIP and PBX Problems?

Big Questions First

The massive transition we’re talking about here is the change from IPv4 to IPv6, or the change from the old Internet Protocol (IP) system to its newest iteration. Int

Now, the old IP is known as the Internet Protocol Version 4, or the IPv4, and it’s the set of rules that have been in place providing the scaffolding for the Internet since day 1. IPv4 works really well, as most internet communication still occurs using its rules, but we’re now transitioning to the new version of the Internet Protocol, a new set of scaffolding known as the Internet Protocol version 6 or IPv6.ernet Protocol lies at the heart of how the Internet works; it defines the way data packets transfer from one connected device to another over various bundles of equipment, cables and wireless signals. The Internet Protocol outlines the rules for how these data packets are labelled, how they are located, and how they are routed over the web.

So basically we’re in the middle, or more accurately at the beginning, of transition from IPv4 to IPv6.

Why Make the Change?

The primary reason IPv6 is being implemented is the fact IPv4 is running out of IP addresses (IPs). In other words, the Internet is running out of addresses that can be assigned to various devices while the number of connectable devices grows each day. IPv6, if implemented worldwide, will take us from 4 billion to over 340 trillion addresses, allowing the Internet to continue to grow indefinitely in the foreseeable future.

Going from 32-bit to 128-bit numbering is not the only benefit of the IPv6. It seems like it’s going to make VoIP work even better than it currently operates. A big reason IPv6 will be a real boon to the world of VoIP lies in the fact the new protocol includes numerous features that make routing and packet processing more efficient, reduces bandwidth consumption, and, finally offers improved Quality of Service (QoS).

Alleviating Address Crunch

The demand for Internet address space exceeded the number of available addresses years ago. To help deal with this problem in IPv4, a complex solution was introduced that allows reuse of non-routable addresses by mapping them on to real addresses. This technique is called Network Address Translation (NAT). In some countries, however, the shortage of addresses is so great that multiple levels of translation are now required. Translation puts demands on networking equipment, complicates routing, requires additional memory and makes firewalls difficult to manage.

Autoconfiguration

Provisioning of all IP devices (including VoIP phones) with IPv6 is much easier than with IPv4. Any network device can allocate its own unique address by appending its layer 2 (MAC) address in EUI format to the 64-bit local link prefix. The address can then be globally routed. There is no need to maintain DHCP server for address management.

Eliminating Network Address Translation

IPv6 handles this issue of open connectivity a lot better by creating a unified address plan that allows any device on the internet to contact any other device quickly and easily. There is no more need for NAT traversal because there is no longer a need for NAT.

You can still set up firewalls to limit access to your network and ensure security, of course, but IPv6 will ensure we get rid of the hardware-based biases that make VoIP tricky, biases that force VoIP technology to trick routers into allowing access to the private IP range our desired phone utilizes.

Improved Routing

IPv6 allows for better address space aggregation thus reducing the size of routing tables in the routers. While this does not mean that routers will use any less memory with IPv6 (after all we are going from 32-bit to 128-bit addresses), this design allows for more efficient routing tables.

Handling of the Fragmentation

Another benefit is that IPv6 no longer requires routers to perform maximum packet size calculation each step of the way. All of the fragmentation in IPv6 is handled by the sending application. That means that sending device “figures out” what is the maximum amount of data can be sent in a payload over any given path. This approach offers lower latency and fragmentation but may lead to inefficient bandwidth utilization.

Better Packet Handling

Unlike IPv4, IPv6 does not store checksum in the packet header and nowhere along the path does this checksum get recalculated. The checksum was dropped because most Layer 1 and 2 technologies already have their own error detection and handling algorithms built in. With IPv4 packet checksum was recalculated at each hop. Not having to do this recalculation expedites routing.

Bandwidth Savings

IPv6 relies on multicast to take advantage of packet flows (such as voice and video streams). Multicast allows sending streams of data to multiple devices simultaneously thus saving bandwidth. The IPv6 header contains flow label which is used to identify packets belonging to specific flow.

QoS improved

With IPv6 each packet is no longer delivered on “best effort basis” as applications may request handling of the packet in accordance with the application-requested priority level.

Improving Peer-To-Peer Communications

Most forms of communication within the internet are decidedly one way. When you browse a web site your browser grabs the necessary files from a server somewhere. When you send an email you just send the message from the server hosting your email platform to the server hosting your recipient’s email platform. Even some seemingly P2P programs, like chat programs, really just function as small, fast versions of email, with each server sending a one-way stream of data to the other as fast as you can type and hit “enter.” All this one-way online communication basically sends and receives very well, but doesn’t let exterior devices “in” to their network space.

Why isn’t Everyone Going for IPv6?

The largest demand for IPv6 comes from the telecommunications industry. Mobile networks continually grow with more and more devices added to them all the time. Companies like Verizon, Comcast, T-Mobile are running externally accessible IPv6 networks. Another set of leaders in IPv6 deployment are content providers such as Google and Facebook who need to be in position to make their content available over IPv6 before everyone else had decided to switch.

But most other industries are lagging far behind.

Compatibility

Even though IPv6 is being deployed as we speak and IPv6 will ideally represent no gap or change in service or usage for anyone, there are few technical points you need to keep in mind, technical points which will affect you if you don’t consider them. For example, IPv4 and IPv6 are not compatible with each other, which means new internet technology you purchase (including internet telephony technology) needs to support both IPv4 and IPv6, a feature we’re starting to see on pretty much every new piece of VoIP technology coming our way.

Added Latency During Transition from IPv4 to IPv6

While most of the Internet still runs on IPv4, several transition mechanisms have been offered as a pathway to IPv6. One such option is running multiple protocols side by side. Unfortunately the management costs of such networks and added complexity increases significant, again, with no immediately visible benefit.

Another technique is tunnelling of IPv6 packets over IPv4 protocol. This option is even less practical as it adds latency, at least until most communications can occur in native IPv6. But that may take a very long time. According to “Global IPv6 Deployment Progress Report” published by Hurricane Electric on October of 2010, only 14% of all autonomous systems on the Internet used IPv6 in some form.

No matter how you look at it – until IPv4 is still a dominant protocol at most Internet exchanges.

Not All Hosted PBX Platforms Support IPv6

There’s one more aspect of the Internet telephony we need to address when it comes to abandoning IPv4, and that’s the world of IP telephony, where the switch is potentially more problematic. We know that IPv6 offers some really good news for VoIP but it unfortunately offers some potential problems for some older PBX and end point devices. These problems are far from insurmountable but they’re issues hosed PBX service providers and end users need to be aware of.

Many IP PBX systems out there either run entirely or ground themselves in older platforms that operate entirely on IPv4. The main issues is the fact that the official versions of some of these popular open source software haven’t begun to officially support IPv6 until recent and few don’t support it yet to date. This means that if an organization or a hosted PBX service provider is running their operations off the older version of the software they’re heading towards a major upgrade.

A Whole Lot of Work With Little Immediate Demand

You’re probably wondering why many popular open source PBX solutions have been slow to make the jump to supporting IPv6 just yet. That’s a good question to ask, a question with a handful of simple answers. The main reason is because the market hasn’t demanded it yet. Most of the internet and most Internet-enabled devices still operate using IPv4 and while IPv6 is currently being rolled out it’s going to be a little while before the new protocol will offer the web’s basic method of operating.

No Clear ROI for Most SMBs

This lack of immediate demand is compounded by a previously mentioned aspect of IPv6, which is the fact that (if all goes well) the switch away from IPv4 will transpire invisibly. For most SMBs there is no visible benefit to switch because until the protocol is widely adopted, the switch will not result in any awesome and marketable new functions or features for users.

For that same reason the development community and small business have been reluctant to aggressively promote changes to the PBX source code. Changing the code in an existing VoIP PBX platform to make it IPv6 compatible requires a LOT of work- a lot of work that not only needs to be completed but which also needs to be implemented and tested one step at a time before it can be rolled out. Open source PBX systems rely on interoperability with the end point devices from various manufacturers. Testing different implementations of the IP stack is a daunting task especially within a loosely knit open source development community.

Taking the prospect of a huge amount of work into consideration, considering that, in fact, the pool of experienced IPv6 network professionals is much smaller than their IPv4 counterparts and combining it with the above mentioned lack of immediately pressing demand to make the switch and you have a clear understanding of why development community and hosted PBX providers aren’t in a hurry to release IPv6 code.

Now, all this sounds a bit alarming, but it’s important to remember that none of these factors became serious problems just yet as IPv4 to IPv6 conversion is likely to take another 3-4 years and there’s a good chance all the open source PBX software will make the switch to IPv6 compatibility. It’s simply a point that PBX providers who haven’t developed IPv6 compatibility within their platforms need to be aware of, and a point clients need to keep an eye and ear out for when picking their next PBX provider.