Is it possible to reconcile the worlds of Panasonic and Asterisk?
Below, we describe a solution to the problem of connecting multiple offices using hybrid Panasonic-Asterisk * by using Alvis-GW-2E1 gateways.
Original facts of the problem:
Suppose there is city A and city B.
In City A there is a Panasonic KX-TDA200 PBX installation that is connected its internal extensions with the outside world.
Also there is a branch in City B: where Panasonic KX-TDA100 PBX system is installed,which also have extensions and access to the PSTN.
For a long time the connections between branches carried out through E1 PRI, but recently operator become trickery and implemented TDMoP between cities – the result of the implementation, both office Telephone Exchange users began to complain about the quality of communication.
Also, the Company’s management is not against the possibility of cheaper long distance calls and cheaper calls to mobiles.
As an option – E1 link between offices do not exist and a decision for a solution is in progress.
Why TDMoP is bad?
The thing is E1 and PRI are like signalization in particular – channel. The Internet – packet transmission. ISDN PRI (Q.931) developers could not imagine that in the 21st century will come to mind to someone to change the normal links on pseudo-e1 (pseudowire). Packet Transmission protocols TDMoP / CeS / TDMoE imposes its own characteristics: as decent Jitter (mixing) package, the possibility of partial packet loss, problems with the Recovery shreds – especially with long delay links.
Lost packets on the terminal side of the link (TDMoP equipmet) either replaces with duplicates of last successfully received E1 frame, or with a frame filler (usually D5 or FF). Imagine what happens to the PRI signaling link, based on HDLC! Messages are lost, calls are hanging.. All this problems are further complicated by incomplete implementation of the PRI stack in modern PBX and Asterisk-systems! Misconfigured Jitter in the TDMoP tunnel immediately affects all channels, either by wildly increasing delays or by causing gurgling in conversations with a high probability of failure.
The most annoying thing is the subscription line is virtually all transparent: unlike conventional VoIP you can not “ping” through ping or iperf! No way to measure jitter and it is difficult to estimate the loss!
Various alternatives:
1. Purchase native SIPGW
Considering and thinking figured that this scenario is not cheap. The fact that it is necessary to buy expensive licenses to activate SIP trunks. This kind of a policy of the vendor, unfortunately. NAT traversal problem forces to organize VPN very carefully..
2. Connecting to H.323 – again you need to buy SIPGW, impossibility in the base configuration and also make connections with Voip-operators, you need to buy additional codec support, echo cancellation and so forth.. Also overcoming the problems with NAT fall on the shoulders of the users themselves. Debugging H.323 dialects – It is not for the faint hearted. Plus if there is another PBX vendor in city B, LG or Siemens for example, there is an imminent compatibility issues.
3. Using the FXO/FXS gateways and conjunction by analog.
Will work, but quality of communication – так себе, there is no way to determine who is calling, the inability to properly organize the multi-channel and forwarding..
Analog, everything is said. The problem is partly solved by Busy chippers, then the problems with the fax turns out and subsequently will bother more …
Our solution:
We propose to replace the E1 link between the cities into local E1 links to the SIP/E1 Alvis-GW-2E1 gateways, and the communication between the cities perform over SIP.
This avoids PRI trouble, in this scheme, there is no more TDMoP!
Synchronization E1 stream – Local or taken from the TDA. No delays in E1.
Problems are solved with echo cancellation using the built-in hardware echo cancellation LEC module in the gateway , compatible with G.168 +, besides supporting the NEC (Networking Echo Canceller).
Possible to use the full E1 stream, in all slots. Do not need to buy any additional licenses, and SIPGW indeed is not needed. In necessity of bandwidth saving- there are G.729/G.723 codecs, and also there is an AMR-NB support.
Problems with delays / packet loss / jitter are solved by standard for SIP/RTP help protocol.
– Both parties may be behind NAT: no need to complicate the system via VPN.
– Possible to consolidate this solution with VoIP operator connections to reduce the cost of international calls.
– With the help of Asterisk you can choose multiple ways of dialing depending on various factors
You can create a variety of advanced redundancy schemes, to provide FailOver solutions.
In this scheme, there is no implementation problems with point-to-point networks: but also there is no implementation problems with distributed branch networks with the possibility of communication “everyone with everyone.”
Network Monitoring
In addition, there is a possibility of centralized monitoring of the entire network via Web interface through the Alvis Distributed Network tab:
Plus gateways have the SNMP support: except for CPU and available memory you can monitor the status of Layer1 E1 and PRI alarms, receive real-time alarms in case of link and gateway management problems.
Currently add-ons for Web-интерфейса Alvis-GW-2E1 being developed, allows you to use presets for different PBX systems automatically: in this example, in particular for connection to KX-TDA. You do not need to repeat the mistakes of others, simply implement the standard settings! Set up does not requires any knowledge of Linux or Asterisk! Web-based interface simplifies the process a lot.
Security Issues
Currently it is not a secret that various hackers, scans and hacks various VoIP systems, especially SIP. Since we do not supply the equipment on the basis of sold-and-forget, we conduct security audits and minimize the problems of our customers against the part of many thousands of phone bills. The whip must slash on the vine!
Also, for us it was a revelation to learn that some TDA firmwares have vulnerabilities over PRI. Our equipment may be used in PRI-PRI calls as a filter. We have developed a specialized PRI filtration system for some of our clients. We are currently exploring the possibility of porting it for public use.
Critical applications
Also, for particular applications, there are special solutions: Redundancy, possibility to order an Industrial-version of the equipment with extended temperature range, and also there are specialized software tools: such as Built-in E1/T1 analyzer and ISDN PRI Monitor. Also, all of our equipment contains a built-in hardware Watchdog timer, that allows you to restore the system in case of failures.
GarantPlus may allow free evaluation of this solution! You heard right! Full solution. We not only give the equipment itself, but also help to set up as part of E1 and SIP trunk.
Additionally, if you wish, we can dock with different VoIP operators – many have developed excellent partnerships. Particularly worth mentioning МТТ!
Not a problem for the solutions and upgrades and expansion: since the system – modular and external, there is no limit for the number of links.
The critical importance should be noted of using the normal networking equipment such as switches, routers and ethernet. Saving on them, will after cost significant money and human resources for troubleshooting! Avoid cheap routers to activate the SIP ALG!
In addition, we can note that our company is famous for its Technical Support: for 10 years practically we have no thumbs-down for our equipment from customers , we are not looking for super-profits – for us it is more important to build long term partnerships B2B!
During the warranty period (5 years), we are committed to resolve all technical issues, including compatibility issues. And in the post-warranty period, we always help our clients with good advice!
Good luck in reconciling the worlds of Panasonic and Asterisk !!
For a hybrid solution – the future!