The Gigabit Networking laboratory at IIT Bombay has been instrumental in two key technology developments in the telecom arena: the CESR and the TCC.
The CESRs: A telecommunication technology called Carrier Ethernet Switch Routers or CESRs was developed by the Gigabit Networking Laboratory at IIT Bombay. The goal of CESRs was to facilitate telecom service providers to move large chunks of data through its network by acting upon data headers at various levels of service offerings. The first indigenously conceptualized, designed, fabricated and patented CESRs were rolled out in Spring of 2011, and the launch customer for the CESRs was MTNL. MTNL built two data-centers, in Mumbai commissioned in May 2011 that used the CESRs. The data-centers housing financial brokers, websites, surveillance traffic were commissioned in May 2011. IIT Bombay and ECIL (a Department of Atomic Energy PSU), signed an MOU for technology transfer of a telecommunication equipment that as developed by IIT Bombay. ECIL would manufacture CESRs. Three products were built in the CESR family – a small box for your home/office environment that has 8 Ethernet copper ports a 2 or 4 Gigabit Ethernet Fiber ports with scalability built in, a metropolitan network aggregator with 10x1 Gigabit Ethernet ports and 2x10-Gigabit Ethernet ports and a core router. The aggregators’ cornerstone was the 1-microsecond port-to-port interconnection latency and for its size of being a 60Gbps duplex cross-connect, consuming only 28Watts of power. The core router which in its basic configuration supported 4x10Gbps core ports and 8x1Gbps edge ports. The ports in this box supported a technology called OTN or optical transport network that facilitated wavelength division multiplexing of channels and reach up to 1000 km without signal regeneration. This box also had a low latency of 3-5 microseconds and an energy consumption of about 65 Watts. A software or network management system was designed in compliance with the concepts of software defined networks or SDNs, whereby a user could remotely configure network parameters to his taste and set up authenticated services that were configured to meet his requirement. Services could be set up based on IPv4, IPv6, port, TAG, VLAN, MAC from 1Mbps to 100Gbps. The commercial offerings were manufactured by ECIL and had close to 100,000 lines of RTL code 64000 lines of Java code completely designed and developed by IIT Bombay. PCBs between 14 and 22 layers were designed and included high-speed traces for up to 11.1 Gbps line-rates. In addition to the code, the boxes had between 800 and 1500 components. Inventory management and ensuring that when the technology became a product, we had to create a sustainable supply chain were some non-technical challenges. Independent testing and quality assurance were conducted by a team from BARC. A key highlight of the CESR technology was its clocking of a mere 1 microsecond port-to-port latency across 3 layers of the networking stack. The two data-centers in Worli and Belapur work on 56 of our CESRs since May 2011.
The TCC: A 1 Tbps capable SDN cross connect (called TCC) developed under an MOU between DRDO and the Gigabit Networking lab at IIT Bombay, is the first of its kind telecom class box to be indigenously developed in the country. The SDN box is a chassis-based telecom solution that has a 1Tbps capable backplane and multiple input output cards. In a typical scenario, the TCC can be deployed as a metropolitan or a core networking solution whereby it can process several million packets per second performing switching, routing and transport capabilities at carrier-class benchmarks.
Deployment scenarios: The TCC can be deployed to create a pan-India network leading to a multi-terabit backbone. The advantage of the TCC is that no other equipment is needed as part of the backbone – no extra switches, routers or optical gear is needed for performing Layer 0 – Layer 4 functionality in this network. Another deployment scenario is that of providing seamless connectivity in metro networks, such as joining several labs or enterprise points of presence within a city to each other. A third deployment is in the HPC space where multiple top of the rack switches in Ethernet mode can be connected to the TCC, thus facilitating intra HPC or intra data-center connectivity. We envisage that the TCC can be used in all the scenarios such as a pan-India network across strategic and tactical sites; a network within a city connecting multiple sites and within data-centers. Apart from a direct cost-advantage of the TCC as compared to other products, the TCC is completely designed and developed in India – it has no switching/forwarding/routing ASICs, implying that all the code is written by Indians in India with the IPR resting jointly between DRDO and IIT Bombay. The TCC thus is a truly Make in India effort and can be a perfect fit to the Digital India vision. Smaller versions of the TCC are being built leading to enterprise and intra government office connectivity as well. The TCC aims to be a best in its class switching, routing and transport fabric with 5 micro-seconds port to port latency – better than any product in its class.
Architecture: The architecture of the TCC consists of a backplane and multiple IO cards. The IO cards can at this time support multiple protocols such as IPv4, IPv6, VLANs, MPLS LSPs and OTN. Further the IO ports can be programmed by the user to send packets on a particular wavelength resulting in Dense Wavelength Division Multiplexing or DWDM. The IO cards are all stackable and can lead to a multi-terabit switching and routing system. A management complex is developed that consists of a software system that controls TCCs. The drag-and-drop based management/control GUI facilitates service creation, network performance monitoring, and provisioning – all at the click of a mouse to a readily authenticated user.
Services can be set up as point-to-point or multipoint-to-multipoint (MP2MP) services from layer 0 to layer 3. These services can be of granularity 1Mbps all the way to 100Gbps. The platform uses concatenation techniques to multiplex several services while guaranteeing a bound on latencies. A key feature of the platform is that each service is provided with one or more protection routes that are edge and node disjoint. In the event of a failure the traffic on the main working path is seamlessly and within 50ms shifted to an alternate protection path. If the protection path fails, another protection path is ready and waiting and hence at any time the service continues to be ON.
Carrier-class performance: A key attribute of this platform is that it is carrier-class i.e. it is built with robustness at the service, card, node, chip, and management system level. In particular, the TCC has service availability features that are absolutely the best in its class – we provide 30-50ms protection on a per-service basis i.e. whether a service is provisioned as a 10Mbps LSP or a 100Gbps lambda, the protection time is less than 50ms. For each service the TCC provides diagnostics, such as per-service end-to-end latency, jitter, packet loss and deviations which can be programmed as per the user needs.
Software Defined Network adherence: One of the key features of the TCC is that it is the first box of this size that uses the SDN philosophy in its entirety in the country. No other vendor has so far built a completely SDN capable box which boasts of telecom grade gear and a 1Tbps fabric in India. Separating the data or forwarding plane from the control plane is part of the SDN philosophy. A standalone controller is used for programming the network to whatever are the users’ needs such as service set up, bandwidth calendaring (changing bandwidth on demand to a location), allocating higher security levels to a location, facilitating virtual networks embedded on a single physical infrastructure and multiple protocol support on a single platform.