Chat with us, powered by LiveChat

Reliability of Voice over IP

By September 1, 2016 Voip

The question is often asked of us of ‘how reliable is Voice over IP or IP Telephony’. In context, the question is usually around the risk associated with replacing a good old reliable PSTN or Analogue line at a retail location and replacing it with a VoIP solution.

We thought we would answer it here, seems like a good place to do it.

The fundamental benefit of what we today call the Plain old telephone service (POTS). POTS is the voice-grade telephone service that remains the basic form of residential and small business service connection to the telephone network in many parts of the world.

The name is a retronym, and is a reflection of the telephone service still available after the advent of more advanced forms of telephony such as ISDN, mobile phones and VoIP. POTS has been available almost since the introduction of the public telephone system in the late 19th century, in a form mostly unchanged to the normal user despite the introduction of Touch-Tone dialing, electronic telephone exchanges and fiber-optic communication into the public switched telephone network (PSTN).

So, simply put, it is old, it works, not much has happened to innovate on it.

Enter VoIP, in our context, it is pretty simple. Our VoIP offering is really proven digital voice transmission delivered within our core via high speed, private, dark fibre connections. They are however, delivered via our MPLS cloud into a customer location via our private and secure network. We provide Quality of Service or QoS markings to ensure the the voice traffic is reliably delivered and appropriately prioritized over other traffic that may be going over the same link.

So far so good, but why from time to time do we hear of VoIP horror stories? Quick answer, if you look long and hard you will always find a good horror story. My favorites are examples where hardware manufacturers, who in a competitive scenario, will compare tales of terror of each parties installations. Example, an example where a Cisco Call-manager implementation failed, or where a Mitel 3300 failed or a Nortel implementation. In these examples, you will find successful installations, yet you can find failed installations.

The difference really comes right down to the following:

– Who is the provider
– Where is their equipment relative to where the calls are being routed
– What was their implementation plan
– What network was it running on
– Are they skilled in delivering this service
– What is their service and support responsiveness
– Can they, or their network scale

Let us explain further:

– Who is the provider

My favorite horror story was with a small provider who operated a VoIP network out of basement of a home. My second being a provider in the Niagara Region who operated an open source PBX off a low grade commercial internet service with DID’s routed from Montreal and New York over the public internet.

There are big guys and there are little guys. VoIP is really a ‘size does not matter business’ but what will always matter is the magnitude of investment of assets and network. Having a highly redundant call control platform and session border controllers (SBC’s) are a good start. But having an entire network that considers the importance of improving the legacy PSTN 5-9′s in reliability is really more ideal.

– Where is their equipment relative to where the calls are being routed

We eluded to this one in our point above. If the provider call control platform is located in Toronto, and the customer is in Calgary, and the originating call is coming from New York – you will quickly see that the inefficiency of distance and design will almost immediately present itself. With Frontier, we have built our call control platform and SBC’s to be regional in nature, thus reducing or eliminating the ‘hairpin’ effect and ultimately reducing latency – which is the largest deterrent to VoIP success. Hairpin implies sending a call back to its source. One of the situations in which this happens is when a call fails to be routed successfully to the destination IP address.

– What was their implementation plan

We have seen examples of failed VoIP projects where the implementation plan just did not exist. Highly complex routing scenario’s and VPN configurations within the customer network are large causes for VoIP failure. Also, a VoIP ready phone system or PBX may require additional complex programming to ‘tweak’ the service. If the VoIP provider and the PBX vendor are not working together, VoIP success will be difficult.

– What network was it running on

There is a strong case for Frontier’s success in the VoIP space which is that we run our VoIP services ‘exclusively’ within our own networks. If you want our VoIP product, we will build a Voice WAN for you to run it on. This ensures success, eliminates finger pointing and still allows us to demonstrate versatility and savings.

Most VoIP, is actually sold and provided to customer locations over the actual Internet vs. a private network. It is an ‘ok’ way to do it. The problem comes down to consistency. There is a funny and effective adage that states ‘you can have good, reliable or cheap – pick two!’. When routing over the public internet, you are in fact getting one, ‘cheap’. There is a better way – go private.

– Are they skilled in delivering this service

There are vendors with specialty in legacy voice or the sale of PBX phone systems, and there are vendors with a specialty towards the sale of network services or LAN gear. Most, even today, don’t posses both skills. Work with a vendor that does. In many cases, despite a PBX having inherent SIP trunking capabilities, the use of a ‘gateway’ provides for a definitive point of demarcation between the VoIP and SIP trunks and the customer premise PBX.

– What is their service and support responsiveness

If there is an issue, can the vendor trouble shoot the problem end-to-end? If not, the finger pointing with ensue. Or worse, it just will go unresolved.

– Can they, or their network scale

VoIP congestion in a service provider network does not always happen at the customer network.

Deliver sufficient throughput: The amount of bandwidth required for an enterprise-quality VoIP business solution depends on the number of simultaneous calls, the signaling overhead, and the voice encoding scheme used in the IP handset or softphone.

The International Telecommunications Union (ITU) G.711 codec is commonly used in LAN deployments where LAN bandwidth is plentiful. With G.711 and RTP header compression, each call requires 82 Kbps. So, if you have a 1.5 mbps link, you need to be mindful of the total number of calls that you can effectively pass within your network. Similarly, the service provider also requires the same path engineering. In this case, they must provide path studies that will actually ‘anticipate’ busy hour call attempts to ensure that suitable capacity exists.

How successful will your implementation be?

The diagram below summarizes the relationship between mean opinion score (MOS) and three primary impairments of packet loss, latency and jitter. Toll quality MOS is considered to be 4.0 or higher. Subscribers begin to hear noticeable degradation to speech quality when MOS falls to 3.5 or lower.

At Frontier, our objective to to engineer the total end-to-end solution within the ‘Good’ range below on the MOS table.

About Frontier Networks

ORANGE_03

Frontier Networks Inc. is headquartered in Toronto, Ontario and provides Broadband Internet or MPLS, Voice lines or Cloud PBX (a replacement to old phone systems), cloud servers , colocation and our new Physical Security offering to Canadian Retail and Multi Site customers who demand world-wide coverage from a ‘new’ network. “We like to do traditional things in a non-traditional way”.  Frontier has built a network that connects to other networks. Think of them like a large ‘backbone’ of interconnected networks. They connect to every phone company, cable company, wireless and hydro/utelco in Canada and the US through a series of well-connected Points of Presence (POPs). Simply put ‘we don’t suck’.

See more at: http://www.frontiernetworks.ca/blog

 

1-866-833-2323 sales@frontiernetworks.ca