A blog post and video demo from our friend and collaborator Murray shows how the BlueVia service can implement an advertising-funded calling model.
Allowing users to stack-up call minutes by volunteering to listen to adverts is a model that has been used in South America, and there is still a lot of industry effort involved in connecting mobile to ad-funding in any number of ways using data/web, voice, SMS.
Murray points out that in the BlueVia model, there’s no need to switch the call out to another voice server and have the complexity of managing 2 call-legs, and using up extra trunks and so forth. Within the O2 network it makes more sense to use IN to change the processing of the call; the user’s account is marked with a special flag (in the HLR) so that whenever a call takes place the “intelligence” in the network allows the calls to always be processed by a BlueVia application. In this case the logic allows audio play out for the ad, followed by proceeding to connect the two-way voice channel. Very neat compared to some solutions.
By the way, Voxygen’s cloud platform is a contributing piece of technology to this demo and Voxygen also developed the voice APIs for BlueVia, hence the mention of Voxygen in Murray’s post. RESTful APIs for voice and conference control is one of Voxygen’s areas of expertise.
Posted in Mobile, Voxygen Tech
We’re thrilled that our research paper “Implementing a Superwideband Codec for Smartphone VoIP Services” has been accepted for the European Wireless Conference, hosted at Surrey University in April. The paper concerns work that we did in taking Google’s iSAC wideband speech codec and implementing it on the iPhone as part of the Voxygen Unify VoIP framework. Our Senior Technical Architect, Martyn Davies, collaborated with the Wireless Multimedia & Networking Research Group at Kingston University, which is led by Professor Christos Politis.
You may know that iSAC underpins the functioning of WebRTC: it provides a high quality voice channel to interconnect devices, such as web browsers, for multimedia communication. In this case, we took iSAC out of its normal WebRTC setting and used it to provide high-definition voice for iPad, iPhone and iPod Touch. We also used equipment from Malden Electronics to do objective tests on speech quality. We measured the performance against the Speex codec, recording test results using PESQ and POLQA quality measurement systems. We also did subjective testing against FreeSWITCH, which makes an excellent HD-voice testbed in addition to its many other great features.
So thanks go to our collaborators at Kingston University, Malden Electronics and the many others that provided advice and help including Google, Opticom (the POLQA people) and the FreeSWITCH team. Martyn will be attending EWC 2013 to present the paper.
Posted in Events, Mobile, Voxygen Tech
Our CEO, Dean Elwood, along with executives from Facebook and the GSMA, will be on tomorrow’s judging panel for the 2013 University Mobile Challenge, held in Hall 8 at the Mobile World Congress in Barcelona.
Universities from across the World have competed to create solutions and execute a mobile business idea. The finalists will be judged on a number of criteria including the commercial appeal of the product, and the quality of user experience.
Dean and other judges can be found in Hall 8, lower level theatre E from around 11.30 AM tomorrow, 28th Feb. If you’re at the Fira, do drop in.
Posted in Events, Mobile
Dean and Christian are in Barcelona this week at the 2013 Mobile World Congress. You can find them with our partners at Telefónica. Here’s a snap from the stand in Hall 3, showing the Azul conferencing service, one of the demonstrations:
Azul is an Enterprise focussed VoIP application that runs on Android and iOS.
If you’re at the MWC yourself, why not pop round and say hello. The guys would be happy to speak to you and give you a demo.
Posted in Events, Mobile
← Older posts
Newer posts →
Fujitsu are in the news this week talking about improvements to TCP/IP so that network congestion doesn’t create a consequent collapse in performance as the applications attempt to re-send data that hasn’t been correctly received. There is actually a long history of research into this area, and “congestion collapse” in TCP has had some of the best minds in computer science experiment with different solutions.
It is a particular concern now that so many devices are wireless: where fixed networks tend to suffer from congestion that causes delay to the packets, wireless tends to corrupt more packets, which are then dropped. The congestion control algorithm in TCP responds to packet loss by going into a “slow-down” mode that harms performance. The combination of packet loss and long confirmation time (for example in satelllite links) can kill TCP performance.
A number of different approaches can be tried, including selective acknowledgements (i.e. re-sending only the specific packets lost), forward error correction (adding much more coding to allow reconstruction of lost data) and spoofing (essentially telling lies to local applications about when data was acknowledged). Some enhancements are already in TCP itself; you may have heard about different variants of TCP (called Reno, New Reno and Tahoe) that improve congestion control. There is also UDP Lite (which we’ve blogged about before) that allows bit corruption to be an application problem rather than a network one. This is a good solution for media (voice or video) where the application understands the format and significance of the data, and can often intelligently repair damaged data. A number of codecs (e.g AMR, H.264) can directly take advantage of this.
It’s not very clear what Fujitsu are suggesting, but of course enhancements to TCP or UDP really depend on you installing the same software on all clients and servers concerned. The great power of TCP is in its universality; it’s unquestionably the most successful network protocol ever because exactly the same protocol is in everything from data centres to mobile phones and wireless sensors.
Posted in Mobile, Multimedia