ARITARI
VOICE OPTIMISATION
Routing and Technology
OVERVIEW
Aritari is the leader in Voice Optimisation and provides a unique and powerful range of capabilities that measurably improve the quality of voice calls while reducing the network bandwidth requirements for delivery.
Our software helps voice deal with the challenges of the Internet namely packet loss and high latency but also
Simplify Voice Routing
One of the largest challenges to voice delivery today relates to complex routing due to a separate voice and data provider and resulting support and quality challenges. Some of the main challenges include:
Protect Voice Quality
Aritari has been designed to improve voice quality in the toughest bandwidth conditions. Our customers have thousands of branches with limited
bandwidth or who struggle to get consistent voice quality from the Internet to their chosen VOIP provider.
Aritari delivers two specific capabilities to address poor network conditions including:
Packet Loss Correction is based on the Aritari proprietary VPN tunnel (not TCP IP) which is more efficient at delivering lost packets in the network than TCP IP. This ensures that when congestion occurs the network is able to resend and accept lost packets faster, improving voice quality.
Reduction in bandwidth used
Aritari reduces the bandwidth requirements to deliver your voice by up to 90% over traditional TCP IP networks. This allows customers with large branch networks to deliver strong quality voice minutes over limited asymmetrical links like ADSL.
An example of this is reducing a VOIP call of 86kbps to 8.36kbps.
Aritari helps customers manage and reduce network and bandwidth cost without compromising on voice quality.
VOIP Gateways
Aritari VOIP Gateways can be deployed to help clients aggregate their voice services to a central DC or to a public VOIP provider network anywhere in the world.
The VOIP Gateway extends the benefits of Aritari Voice Optimisation all the way to your VOIP provider or selected network point to ensure that you get the best network performance available.
Aritari also provide public managed VOIP Gateways in major locations in the world including the US and Europe.
Additional Features Included
VOIP Optimisation from Aritari includes the following additional technology:
ARITARI TECHNICAL OVERVIEW
Technical Challenges – SD Voice
This section describes some of the current challenges facing organisations and VOIP providers in delivering quality voice over the Internet.
Main barriers to deploying VoIP include:
The features of Aritari SD Voice which eliminate or significantly reduce these barriers:
Bandwidth Consumption and Router Load
The essential portion of a voice packet is normally very small. For example, a 20mS G.729 encoded voice packet contains only 20 bytes of useful data, transmitted at a rate of 50 packets per second.
This results in:
50 x 20 x 8 bits per second = 8000 bits per second (8kbit/s).
To be able to carry this packet across an IP network you also need some additional data.
Adding all of this to the standard 20mS G.729 packet our data rate now becomes:
(20 + 20 + 8 + 12) x 50 x 8 = 24 000 (24kbit/s)
Now a voice call is using three times more bandwidth than it needs to. In addition, at layer 2, or the network layer, there is more information required to transfer data packets across physical links.
At best, this link will be Ethernet, which “only” adds another 14 bytes for every packet (making our G.729 voice channel use 29.6kbits/sec.), however ATM based technologies such as most broadband links are less efficient. These networks use fixed sized “cells” of data to carry information, and typically these cells are 53 bytes in length. The G.729 packet above would require two such cells, and now have:
2 x 53 x 50 x 8 = 42400 (42kbit/s)
Therefore each G.729 voice channel that is carried across a broadband link consumes
42.4kbit/s of bandwidth, only 8kbit/s being used for the information that we really want to be transmitted.
How Aritari SD Voice solves this problem:
The information that is added to a voice packet to transmit it across a network is mostly superfluous. This is because all of it is either fixed for the entire duration of the stream, (such as destination address and port), or can be worked out from previous packets, (such as sequence numbers and time stamps). SD Voice creates a VPN tunnel between two SD Voice enabled devices. Within this tunnel, voice packets from multiple channels are combined into a single data stream, which has several effects:
VOICE OPTIMISATION
Routing and Technology
OVERVIEW
Aritari is the leader in Voice Optimisation and provides a unique and powerful range of capabilities that measurably improve the quality of voice calls while reducing the network bandwidth requirements for delivery.
Our software helps voice deal with the challenges of the Internet namely packet loss and high latency but also
- Simplifies Routing
- Protects Voice Quality
- Delivers up to 90% Bandwidth usage reduction
Simplify Voice Routing
One of the largest challenges to voice delivery today relates to complex routing due to a separate voice and data provider and resulting support and quality challenges. Some of the main challenges include:
- Poor voice quality over data network
- Poor voice quality of Internet
- Voice routing issues between voice and data networks
Protect Voice Quality
Aritari has been designed to improve voice quality in the toughest bandwidth conditions. Our customers have thousands of branches with limited
bandwidth or who struggle to get consistent voice quality from the Internet to their chosen VOIP provider.
Aritari delivers two specific capabilities to address poor network conditions including:
- Forward Error Correction
- Packet Loss Correction
Packet Loss Correction is based on the Aritari proprietary VPN tunnel (not TCP IP) which is more efficient at delivering lost packets in the network than TCP IP. This ensures that when congestion occurs the network is able to resend and accept lost packets faster, improving voice quality.
Reduction in bandwidth used
Aritari reduces the bandwidth requirements to deliver your voice by up to 90% over traditional TCP IP networks. This allows customers with large branch networks to deliver strong quality voice minutes over limited asymmetrical links like ADSL.
An example of this is reducing a VOIP call of 86kbps to 8.36kbps.
Aritari helps customers manage and reduce network and bandwidth cost without compromising on voice quality.
VOIP Gateways
Aritari VOIP Gateways can be deployed to help clients aggregate their voice services to a central DC or to a public VOIP provider network anywhere in the world.
The VOIP Gateway extends the benefits of Aritari Voice Optimisation all the way to your VOIP provider or selected network point to ensure that you get the best network performance available.
Aritari also provide public managed VOIP Gateways in major locations in the world including the US and Europe.
Additional Features Included
VOIP Optimisation from Aritari includes the following additional technology:
- Quality of service parameters
- Link bonding and bandwidth aggregation
- Transparent failover
ARITARI TECHNICAL OVERVIEW
Technical Challenges – SD Voice
This section describes some of the current challenges facing organisations and VOIP providers in delivering quality voice over the Internet.
Main barriers to deploying VoIP include:
- Excessive bandwidth consumption of even so-called efficient CODECs such as G.729.
- High load on transit routers due to the large number of packets per second involved where there are many calls being carried
- Latency and jitter which arises as the result of larger data packets using the same links (even evident if the only data traffic is session data such as that provided by SIP)
- The cost of high bandwidth WAN links which are needed to solve these issues using traditional methods
- Lack of CODEC support in devices
- Complexities involved for enterprises wishing to deploy VoIP between sites across the public Internet or non-private links
- The cost of providing backup solutions to avoid the WAN link being a single point of failure
The features of Aritari SD Voice which eliminate or significantly reduce these barriers:
- Bandwidth used by voice is reduced by as much as ten times
- Jitter introduced using router queues is reduced to virtually zero
- Classes of data can receive as little as 0.4kbits/s
- Interactive traffic remains responsive
- There is no need to reduce the maximum transmission unit (MTU) of the WAN
- Aritari SD Voice optionally supports real-time and invisible transcoding of G.711 to more efficient CODECs with higher MOS scores than G.729 (the most widely supported low bandwidth CODEC)
- Backup links can be switched to in less than a second and without losing calls in progress
- Multiple links can be combined to both increase bandwidth available and eliminate single points of failure
- Sites can be privately linked across the public Internet
Bandwidth Consumption and Router Load
The essential portion of a voice packet is normally very small. For example, a 20mS G.729 encoded voice packet contains only 20 bytes of useful data, transmitted at a rate of 50 packets per second.
This results in:
50 x 20 x 8 bits per second = 8000 bits per second (8kbit/s).
To be able to carry this packet across an IP network you also need some additional data.
Adding all of this to the standard 20mS G.729 packet our data rate now becomes:
(20 + 20 + 8 + 12) x 50 x 8 = 24 000 (24kbit/s)
Now a voice call is using three times more bandwidth than it needs to. In addition, at layer 2, or the network layer, there is more information required to transfer data packets across physical links.
At best, this link will be Ethernet, which “only” adds another 14 bytes for every packet (making our G.729 voice channel use 29.6kbits/sec.), however ATM based technologies such as most broadband links are less efficient. These networks use fixed sized “cells” of data to carry information, and typically these cells are 53 bytes in length. The G.729 packet above would require two such cells, and now have:
2 x 53 x 50 x 8 = 42400 (42kbit/s)
Therefore each G.729 voice channel that is carried across a broadband link consumes
42.4kbit/s of bandwidth, only 8kbit/s being used for the information that we really want to be transmitted.
How Aritari SD Voice solves this problem:
The information that is added to a voice packet to transmit it across a network is mostly superfluous. This is because all of it is either fixed for the entire duration of the stream, (such as destination address and port), or can be worked out from previous packets, (such as sequence numbers and time stamps). SD Voice creates a VPN tunnel between two SD Voice enabled devices. Within this tunnel, voice packets from multiple channels are combined into a single data stream, which has several effects:
- SD Voice only needs to send one set of IP and UDP headers for each “super-packet” that it sends
- It allows SD Voice to not transmit the superfluous information for each voice channel at all. In fact, the SD Voice system is so efficient that the total overhead for each channel within the stream is 2.287 bits
- The cell padding effect which is present on broadband networks is eliminated.
