Telecom and Mobile Research

Spoken Web

The basic principle of Spoken Web lies in creating a system analogous to the World Wide Web using a technology most of us all have in common - speech. Spoken Web helps people create voice sites using a simple telephone, mobile or landline. The user gets a unique phone number which is analogous to a URL and when other users access this voice site they get to hear the content uploaded there. Interestingly, all these voice sites can be interlinked creating a massive network, which can work like the World Wide Web. Spoken Web has huge potential in the developing regions space, where most of the people are either illiterate or cannot afford a PC but many of them own a mobile phone and can speak in their local language.

Researchers: Sheetal K Agarwal, Jyoti Grover, Anupam Jain, Arun Kumar, Priyanka Manwani, Sougata Mukherjea, Amit A. Nanavati, Monia Puri, Nitendra Rajput, Meghna Singh, Kundan Shrivastava, Saurabh Srivastava, Jerome S White

Social Network Analysis

An analysis of telecom call social network graphs often provides insights for telecom operators to measure the true value of a customer. The telecom call graph analyses are computationally expensive but play a key role in finding parameters beyond bill generated to mark the importance of each customer to the telecom player. SNAzzy analyzes the social networks graphs formed by telephone calls, voice and SMS (text), among the customers of telecom service providers. It performs a global structural analysis of the graph to determine various structural attributes of the call graph such as degree distribution of telecom customers in terms of incoming and outgoing calls (and other attributes). The technology finds structural communities in the call graph such as cliques, stars and dense sub-graphs among others, and computes telecom customers with a risk of churn from the social angle.

Researchers: Kuntal Dey, Ananth Dwivedi, Hemank Lamba, Natwar Modani, Sougata Mukherjea, Seema Nagar, Amit A. Nanavati, Ramasuri Narayanam

BusinessFinder

BusinessFinder is an advanced presence-driven, real-time location based service solution providing consumers the ability to establish connection to business services and individual vendors quickly, whenever and wherever needed. BusinessFinder links business information with network capabilities to build a dynamic real-time repository offering that leverages multiple information channels and communication modalities.

With built-in dynamic appointment scheduling, multi-modal subscription channels, and real-time data-driven matching, BusinessFinder provides enormous opportunities for businesses to effectively use real-time data and drive revenue. Such opportunities range from better work force management to enabling real-time social networking. In an open marketplace, it provides significant opportunities to roll-out advertisement portals for micro-businesses where real-time data is one of the key drivers to route requests.

Researchers: Dipanjan Chakraborty, Koustuv Dasgupta, Sumit Mittal, Sougata Mukherjea, Seema Nagar

Large Scale Presence Virtualization and Federation

Presence has rapidly evolved to become the de-facto method of representing and querying the dynamic status of an individual, device or abstract entity (for example, the number of attendees in a conference call) in a variety of Web- based content provider, enterprise and service provider converged applications.

With the proliferation of presence, an individual's contextual state is increasingly fragmented across different applications and provider domains. Currently, presence- based applications operate in domain-specific silos, unaware of the individual's presence status in other domains. To support increasingly sophisticated, large-scale deployment of presence-based applications -- for example a call-center (Helpdesk) monitoring application interested in the percentage of call- center employees who are available -- we are developing a presence virtualization layer that provides a programmable abstraction by which applications can easily query for their desired collective view of presence, without focusing on the details of individual "presentities".

Our presence virtualization middleware tackles the two important challenges of query flexibility (via a programming model that is expressive enough to support a wide variety of virtualization queries) and scalability (by controlling both the network traffic load and the query processing overhead).

Researchers: Arup Acharya, Nilanjan Banerjee, Dipanjan Chakraborty, Koustuv Dasgupta, Shachi Sharma, Xiping Wang

Telecom and Mobile Application Framework

Our goal is to enable end-to-end lifecycle of applications in the Telecom and Mobile domain, from creation to delivery. We can classify the projects in this area into several major efforts:
a) Creation of Services Utilizing Telecom Functionality: With Telecom market reaching saturation in many geographies and revenues from voice calls decreasing, telecom operators are trying to identify new sources of revenue. For this purpose, these operators can take advantage of their core functionalities like location, call control, etc. by exposing them as services to be composed by developers with third party offerings available over the Web. In this effort, we offer a framework for creating applications exploiting telecom functionality by first providing an abstraction model for encapsulating invocation, coordination and enrichment of the telecom functionalities, and thereafter rendering a service creation environment on top of this model.
Additionally, in coordination with IBM Research - Haifa, we developed a modeling toolkit that uses the above framework to offer modeling based telecom application development environment, more suited for traditional application development.
b) Handling Fragmentation of Platform Interfaces for Mobile Applications: To foster creation of rich mobile applications, popular platforms such as Android, iPhone and Nokia S60 offer extensive middleware support. This includes not only helping developers code and package their application modules in a format suitable for deployment, but also providing "interfaces" to access information on the mobile device (for e.g. user location) and invoke device capabilities (like camera), from within the applications. Although the usage of such platform interfaces leads to richer modules, it requires the developer to deal with application fragmentation arising due to heterogeneity in syntax, semantics and implementation of these interfaces across different platforms. In this effort, we first develop a semantically structured unit to absorb platform interface heterogeneity, and then use it as a building block to develop a middleware "de-fragmentation" layer for mobile applications.
c) Next-Generation Mobile Web Mashups: The evolution of Web browser functionality on mobile devices is the driving force for mobile mashups, whereby content rendered on a device is an amalgamation from multiple Web sources. In this effort, we present a framework that take mashups in this domain to the next level of richness, by incorporating two additional dimensions - (1) capability to integrate native device features, such as location and calendar information, camera, bluetooth, etc., within a mobile Web application, and (2) availability of core telecom network functionality, such as SMS and Third Party Call Control, as services in a converged IP/Web network setup. Furthermore, the framework provides mechanisms to increase portability of mashups across different mobile platforms.
d) Application Store Enablement: Telcos would need to enable a host of third party developers to easily offer applications for the Telco end customer. If these applications and services are to be manually on-boarded in a Telco app store, then the solution will not be scalable. Therefore, a Telco needs a way to automate the on-boarding of applications and services in the Telco’s app store and handle the associated execution dependencies so that we can correctly execute the application on a device or in a cloud. However, since there are so many varied device platforms and services, automating the end-to-end process is a challenge.
To help Telcos overcome this challenge, we are developing an Application Store Enablement Framework that help Telcos automate the task of on-boarding applications and cloud services in an app store.

Researchers: Vikas Agarwal, Rema Ananthanarayanan, Sunil Goyal, Sumit Mittal, Sougata Mukherjea

Enterprise Level Secure Mobile Endpoint Management

The last few years has seen an unprecedent surge in the deployment and use of various kinds of end point devices such as smart-phone carrying human sensors to various smarter planet initiatives such as smart meters for the electric grid, smart traffic and so on. Given this rapid proliferation of end point devices and our growing dependence on them for critical operations such as nuclear plant monitoring,  it is of great importance that we be able to securely and effectively manage these devices remotely and prevent/detect possible attacks. However, there is an enormous  and growing diversity in the capabilities of the end point devices based on their intended usage. For example, smart phones running the Android, iOS, Symbian, Blackberry and Windows Mobile platforms have a lot more processing capabilities which could be exploited for enforcing the security and privacy policies of the end user/organization through cryptographic operations. Furthermore, these devices carry an array of sensors such as cameras, accelerometers, GPS etc. which can be used in a variety of applications starting from securing the devices through identity management and multi-factor authentication to more consumer driven applications such as gaming. On the other end of the spectrum, low power passive RFID tags are used in freight monitoring and do no offer much with regards to processing power. However, the security and protection of these low end devices is also of paramount importance to the enterprise's operations.

In the Secure End Point Management (SEM) project, we are developing a holistic and extensible framework that would enable enterprises to securely manage these remote endpoints. The principal focus is on enabling the enterprises to protect their devices against security threats at various risk-levels and prevent unauthorized actions on the devices such as the copying/deleting of sensitive information. Our framework also handles the mobility of the end points and the security threats that arise thereof from an enterprise perspective such as the loss of a smart phone carrying sensitive data. We aim at providing the enterprise with a uniform way to accomplish critical functions such as secure key management, policy management, provisioning, and asset management that is applicable across various application domains.  The SEM framework smartly handles the challenges of managing a dynamic set of mobile end-points with different communication protocols and security capabilities and enables the enterprise to effectively and remotely enforce their security policies on their managed endpoint devices.

Researchers: Sougata Mukherjea, Shalini Kapoor, Palanivel Kodeswaran and Vikrant Nandakumar

BlueZen: A programmable, mobile context computing framework for System or Community of People

With the sharp rise in smartphone adoption in the recent years and the enhanced sensing capabilities now available on the smartphones, several business domains such as healthcare, transportation, energy, public safety, public services, are actively exploring the paradigm of participatory sensing that promises the prospect of leveraging a system or community of people carrying smartphones for large scale sensing activity. The general idea is to engage a community of people to form a distributed, autonomous network of smartphone sensors that could potentially provide contextually most relevant information in real-time to empower the business applications. Engaging the community passively at the right point in time and space is, however, a non-trivial problem. Most of the community oriented applications that are available today are point applications with limited scope. Smartphone users today manually contribute with the data (often using the sensors on the phone) for individual applications that result in high cognitive load. Automated sensing is notoriously difficult "in the wild" and hence the sensing resources available on the smartphones mostly remain under-utilized today. Besides, an important challenge in participatory sensing is to motivate and encourage the participation of a user, while sustaining it over time.

We are developing the BlueZen framework to solve these problems and enable sustainable development and deployment of community oriented, participatory sensing applications. In particular, BlueZen provides a platform for developing and deploying community oriented applications with the following benefits to the ecosystem of people and the business applications:

• Connects the application and the users who are interested in participatory sensing, effectively, efficiently, and passively, while honouring requirements and restrictions of individual user and application.
• Intelligently manages the mobile sensing infrastructure for different classes of community-oriented applications, targeted towards a diverse range of business domains including smarter cities, smart mobile commerce, healthcare, etc.
• Provides an easy, flexible programming interface for developing community oriented applications leveraging the community of people interested in participatory sensing.

Researchers: Nilanjan Banerjee, Dipanjan Chakraborty, Sumit Mittal

Past Projects

Service Control Layer (SCL)

Telecom operators have started to make significant investments towards evolving their Service Delivery Platforms (SDPs) to next-generation architecture, based on IP technologies and standards such as SIP, SOA and increasingly, Web 2.0. These investments are tightly coupled with demands for network middleware infrastructures that allow for rapid service creation and execution, efficient service orchestration and management, and flexible exposure of core network capabilities as services. With consumers and businesses increasingly relying on Web 2.0 oriented services for information, communications, collaboration and entertainment - the key to this evolution would be the integration of Web/SOA middleware and Telecom middleware infrastructure, thereby creating a converged Web-Telecom services layer. To this end, operators are gradually embracing the IP Multimedia Subsystem (IMS) and looking at developing an ecosystem of IMS/SIP based services that can be mashed up with Web based applications, in innovative ways.

Service delivery is vital for the success of the IMS based approach. Current approaches typically focus on chaining existing services into a new composite service. However, for next-generation service delivery, operators will require greater flexibility around real-time management, control and enhancement of user services at runtime - in a manner that is decoupled from the application logic itself. We observe this need and develop the Service Control Layer (SCL), a middleware technology that sits between the IMS service and network plane, and provides operators with the requisite fine-grained control and enrichment over the delivery of next-generation converged services.

Researchers:Nilanjan Banerjee, Koustuv Dasgupta, Sougata Mukherjea, Seema Nagar. In collaboration with IBM Research - China.