Tag Archive for 'mobile cloud'

How HTML5 advances Mobile Cloud Computing!

When it comes to mobile cloud computing there are several inherent obstacles that currently limits its growth and prevalence. Most of these have to do with restrictions related to mobile networks, namely variable data rates, intermittent connectivity and longer network latencies, compared to fixed broadband. These limitations are described in more details in an interesting article from Alcatel-Lucent called Mobile Cloud Computing Challenges. Although future 4G networks, like pure IP-based LTE (Long Term Evolution) will increase network bandwidth capacity greatly, with a potential downlink rate of 100Mbit/s and uplink of 50Mbit/s, mobile networks will always have their fundamental limitations  arising from radio spectrum limits.

Nevertheless, the analyst/research companies have been busy issuing new  studies that forecast a huge growth in mobile cloud services and applications. According to a study from Juniper Research from earlier this year, the market for mobile apps will grow 88% from 2009 to 2014 (to $9.5 billion). ABI Research forecasts that applications and services will migrate and be provisioned from the mobile cloud, instead of being downloaded and installed on mobile smartphones and tablets/netbooks.

Mobile cloud computing essentially means that the processing and data storage of applications is being performed by a cloud based IT infrastructure rather than on the native handset. Of course many service providers, such as Google and Facebook, provide mobile cloud services already. Normally, services are either fully accessed through a smartphone browser interface or through a hybrid approach by installing a light-client on the handset that performs some intermediate processing and lessens network burden, but otherwise connects and fetches services from the cloud.

While most apps are currently being downloaded and installed locally on the handsets, mobile cloud apps will enable users to run the directly from the cloud. This will bring several implications. For example, new type of mobile cloud phones, that are essentially less powerful smartphones, will start emerging. As discusses in one of my earlier posts, these handsets are specifically oriented toward running mobile cloud apps. Obviously, high-end smartphones will continue to be developed, particularly to manage processing intensive apps, such as games and graphic intensive apps.

However, due to the intrinsic limitations and latency concerns in the mobile network, new technologies and processes are needed to provide acceptable quality of service and user acceptance.

One of the technologies spurred to advance mobile cloud computing services is definitely HTML5, the latest version of the Web´s markup language. But what is it with HTML5 that’s important for mobile cloud computing? Well, there are several important elements:

  • On-device caching – the possibility to store data off-line on the device for further processing, reducing the problem of interruptions in mobile connectivity and discontinuous user experience
  • Canvas and video – making it possible to add graphics and video to a web page without having to install third-party plugins like Flash or Silverlight. When smartphones will generally support HTML5, they provide a powerful way to publish and process media on a page.
  • Geo location – as an additional specification to HTML5, GeoLocation API will utilize smartphones location capabilities for integration with mobile cloud services and apps.
  • Advanced form – mobile devices will benefit from more capable forms than in previous HTML versions, especially as they will reduce the need of downloading Javascript code and cause less round-trips between device and cloud server as the HTML5 browser becomes more capable.

These are some of the important mobile cloud enablers introduced by HMTL5 that should lead to a boost in new mobile cloud services and applications. However, not all smartphone platform and hardware developers have started to support HTML5. Both the Android and iPhone already do, so  it is only a matter of time when more mobile cloud apps start emerging.

Cloud computing and mobile VoIP

Mobile VoIP (mobile Voice-over-IP) can be categorized as a Cloud service if it provisioned through SIP gateways that smartphones running VoIP clients connect to. Juniper Research expects that mVoIP will grow from 15 billion minutes this year to 470 billion minutes by 2015 – that’s multiplying the current number with 31.

It’s been suggested by many, and also hotly debated, that mVoIP is the number one threat that mobile network operators (MNOs) are currently facing – causing a significant pressure on their current business models, i.e. primarily resulting from selling mobile voice minutes. But how exactly will mVoIP be enabled by cloud computing?

Basically, mVoIP can be implemented and used in two different ways. Either a mobile device (smartphone) is used as a SIP client (SIP is a signaling protocol) that uses a data network to send and receive SIP messages and to send/receive RTP voice packets (RTP is a standardized packet format for delivering audio and video over a IP-network) or, a SIP gateway is used to bridge SIP and RTP into the mobile network’s SS7 infrastructure. For supporting mVoIP, high-speed IP-networks are used, including Wi-Fi and 3G/4G mobile networks.

Most smartphones today offer both mobile and Wi-Fi interfaces and can be adjusted to prefer connections to a Wi-Fi network, e.g. a hot-spot, when in range and given adequate bandwidth or service quality. There are basically three scenarios for mVoIP that users can choose from, depending upon the type of device, and device status, they are calling to on the other end:

  1. A smartphone user calls from his mVoIP enabled smartphone over a Wi-Fi network to another wireless device connected to a Wi-Fi, either another mVoIP enabled smartphone or other Wi-Fi enabled device, e.g. Apple iTouch or a PC, etc. The limitations are that devices at both ends need to have a client installed from the same service provider, e.g. Skype or TruPhone. In this case the call is transferred via the Internet and the mobile network is bypassed entirely, meaning the MNO gets no mobile traffic revenues.
  2. Secondly, a user calls from his smartphone’s VoIP client to a regular phone number, either fixed or mobile, anywhere in the world. In this case the caller must have purchased VoIP “minutes” from the service provider, e.g. “Skype Out”. The MNO gets some revenues from the VoIP service provider, e.g. Skype, for the mobile minutes sold to the service provider, most probably in wholesale.
  3. Lastly, a user calls from his smartphone’s VoIP client to another smartphone with a similar or equal client and both smartphones are connected to the mobile network, i.e. 3G network. In this case the MNO only receives some revenues from data services, i.e. no charges for mobile minutes.

From a pure end-user price perspective, and especially for international calls, mVoIP services of this type is win out in most comparison.

However, there are limitations to mVoIP services, especially from the cloud based service providers like Skype, TruPhones and others. These services are not part of the global numbering plan, meaning that the VoIP client does not receive a regular mobile phone number. On the other hand, many telecom operator have implemented their own VoIP platforms, although still not many mVoIP platforms, and integrated to their numbering plan. Those services are mostly for international calling.

With more and more cloud service providers integrating mVoIP into their service offerings, including SaaS providers like Facebook and Google, it seems evident that mVoIP will play a major role in the future. Whether it is going to completely transform the way we use the mobile telephone still remains to be seen.

The emergence of mobile cloud computing

Did yo know that digital media entertainment (DME) traffic currently represented by the iPhone is already 35 times higher than that created by traditional handsets? Many analysts believe that most mobile apps will move from being handset-centric (thick-clients) to the cloud so that mobile terminals will run applications directly from the cloud – especially as apps become larger and feature rich. Today, this is not necessarily the case.

Take iPhone for example – where users purchase and download apps from iTunes AppStore for running them on the phone, although some apps may support back-end processing and data storage as well. Evidently this will change – and why shouldn’t it? Mobile cloud computing is certainly just another method of delivering software applications in a service fashion over the network – in this case wirelessly, e.g. via WiFi and mobile broadband networks such as 3G and WiMax. Mobile network equipment manufacturers and providers are promising that their future Long-Term-Evolution (LTE) platforms will support approximately 170Mbit/s data rate speed (uplink and downlink) in each cell – although in reality the throughput is probably going to be somewhat lower. Obviously the available speed will be dynamically distributed between users within each cell – the more users the less bandwidth for each user.

Users will be able to seamlessly transfer from WiFi networks to mobile operator broadband networks, and vice versa, without necessarily knowing what network he or she is accessing at any particular time. The handset, e.g. smartphone, will make sure that it is selecting and using the most beneficial connection in terms of cost and quality.

Mobile cloud computing

Clearly, this will improve user experience and flexibility and enhance service delivery. Users can access SaaS apps irrelevant of their device, fixed or wireless, PC or smartphone. It should be noted that many analysts, including ABI Research, assert that cloud computing will soon become a disruptive force in the mobile world and, eventually, becoming the dominant way in which mobile applications operate and are delivered.