DANCING NEBULA

DANCING NEBULA
When the gods dance...

Friday, March 30, 2012

LifeNet: About LifeNet

Home About LifeNet Download Organization Updates Blog

Home >> About LifeNet

Introduction
In the wake of major disasters, the failure of existing communications infrastructure and the subsequent lack of an effective communication solution results in increased risk, inefficiencies, damage and casualties. Current options such as satellite communication are expensive and have limited functionality. A robust communication solution should be affordable, easy-to-deploy, require low-to-zero infrastructure, consume little power and facilitate Internet access.

LifeNet is a WiFi-based data communication solution designed for post-disaster scenarios. It is open-source software and designed to run on consumer devices such as laptops, smart-phones and wireless routers. LifeNet is an ad hoc networking platform over which critical software applications including chat, voice messaging, MIS systems, etc. can be easily deployed. LifeNet can grow incrementally, is robust to node failures and enables Internet sharing. A novel multi-path ad-hoc routing protocol present at its core, enables LifeNet to achieve these features.

Problems in existing communication technologies
The primary drawback of existing communication technologies such as cellular networks, WiFi networks, etc. is that reliability is not built into their design. They are designed and engineered to be efficient; reliability and fault-tolerance are secondary. Reliability is typically traded off for ''performance at optimal cost''. Their designs often evolve into single-point failure systems making them vulnerable to disasters.

Figure 1 - Typical Cellular Network


Figure 1 is the schematic of a typical cellular network. Various types of cellular networks that we see today, such as GSM, CDMA, 3G, 4G, etc. all have a similar architecture. Mobile Switching Centre (MSC), Base Station Controller (BSC), Base Transceiver Station (BTS) and cell phones are the main blocks of the architecture. The wireless network is divided into a number of cells, each defined by a radio frequency (RF) radiation pattern from a respective BTS antenna. Every cell contains several cell phones that directly communicate only to the BTS. The BSCs mediate the communication between BTSs and MSC. MSC is the main node of the network, that connects a network under itself to outside networks (their respective MSCs) and handles all the required routing and packet-switching. Since this architecture provides a clear functional hierarchy and optimal communication, technologies have evolved over the years but this architecture has persisted.

Since all the key components, such as MSCs, BSCs or BTSs function hierarchically, and together they have a single point of failure. MSCs are present at the root of the hierarchy. Failure of an MSC leads to the failure of the entire network (Figure 2(a)). BSC's failure leads to the failure of the network below it in the hierarchy (Figure 2(b)). When any BTS fails, then the communication in its entire cell is hampered (Figure 2(c)). Although this architecture provides optimal performance, it lacks fault-tolerance and reliability and hence remains vulnerable to disasters.

Figure 2: (a) - MSC failure, (b) - BSC failure, (c) - BTS failure

WiFi is another type of wireless network that we use everyday. Figure 3(a) is the schematic of a typical WiFi network. WiFi network usually consists of end-user devices such as laptops, smart-phones, etc. associated to a WiFi router. WiFi router interfaces the Internet connection (usually through a modem and a cable) with the WiFi network. It handles the required routing and packet switching.

Figure 3: (a) - Typical WiFi Network, (b) - WiFi Router failure

WiFi is also a single point failure system. As Figure 3(b) shows, if the WiFi router fails, the rest of the end-user devices cannot communicate with each other in spite of having the capability to do so. Most WiFi drivers come with a built-in mode of communication called as ad hoc mode. Although this functionality is useful, it does not have any routing capability.

LifeNet: A solution
In scenarios such as communication in disaster relief, wireless sensor networks, etc. reliability of connectivity is more important and bandwidth requirements are not too stringent. It is critical to establish a baseline wireless channel over which users can communicate and coordinate their on-field activities. The communication solution should be rapidly deployable, self-powered, robust to failures, locally maintainable and extremely easy to use.

I. Ad hoc communication
The first and the most important design decision that we made was to adopt the paradigm of ad hoc communication. All devices on a LifeNet network are considered peers without a hierarchy.

The absence of any functional hierarchy across the system prevents it from becoming a single point failure system. Figure 4 depicts the idea of ad hoc networking. Different end-user devices such as laptops, smart-phones and routers communicate with each other in an ad hoc fashion without any centrally important governing device. Two devices that are close by can communicate directly with each other, whereas communication between far off devices can be relayed by intermediate nodes. This design is the basis behind our argument that LifeNet is fault-tolerant by design. The key challenge that LifeNet addresses is achieving a practical trade-off between the mutually conflicting goals of reliability (fault-tolerance), efficiency and usability.

II. Use of commodity hardware
For higher acceptance levels, a new technology should seamlessly plug into the existing technologies around it. This holds true for both the technologies that the new technology consumes and the technologies that consume the new technology. Re-inventing the wheel may sometimes seem an ideal approach in theory, but it seldom works in practice. We understand this fact and have designed LifeNet accordingly.

Figure 5(a) shows the standard layered wireless networking stack. The bottom two layers (Physical and MAC layer) are dependent on the radio frequency (RF) communication technology used and are hence hardware dependent. The higher layers are pure software components of the operating system and are completely independent of the hardware.

LifeNet is designed as a new pluggable software layer to sit between the MAC and Network layers. It does not require any modifications to the layers that consume it (Network and higher layers) and the layer that it consumes (MAC layer). This design decision allows LifeNet to be interoperable with various popular hardware platforms such as laptops, routers and smart-phones. Along with devices such as routers, LifeNet makes consumer devices like laptops and smart phones capable of forming ad hoc networks. This design decision also simplifies the porting effort onto different RF communication technologies such as different flavours of WiFi 802.11 a/b/g/n, WiMax, etc.

III. Fault-tolerant multipath routing

As shown in Figure 6(a) LifeNet is designed to use a multipath routing protocol for communication between devices. This protocol, called 'Flexible Routing', lies at the heart of LifeNet and makes it useful for transient environments. By transience we refer to devices moving in the network, device failures, dynamic network traffic conditions, changing physical obstructions, interference, etc. Disaster relief operations, wireless sensor networks are all highly transient environments. Since the routing protocol, called as 'Flexible Routing' is capable of delivering packets under varying degrees of transience, it makes LifeNet a promising solution for transient environments. Figures 6(b) and 6(c) show the importance of multipath routing in handling node failures.

IV. Self-discovery and incremental growth
The current software implementation of LifeNet is based on WiFi (802.11 a/b/g) and runs on commodity devices including laptops, smart-phones and WiFi routers. LifeNet software has a straight-forward installation procedure and does not require the user to perform tedious configuration settings. Once switched ON, the self-discovery mechanisms built into LifeNet automatically discover other users on the network and present a global view of the network to each user. Users can then initiate communication to an individual user (unicast) or a set of users (multicast).

LifeNet does not impose any constraints on network topology. The network can grow and take any shape as new users incrementally join the network.

V. Minimal infrastructure requirements
Traditional wireless communication systems are infrastructure-based. Consider cellular network as an example. Inside a cell, BTS is the routing authority and all end-user devices i.e. cell phones directly communicate only with the BTS. The average radius of a cell is supposed to be somewhere around 1-2 Kms. However, it can vary depending upon the user-density. Thus, capability for communication with every active cell phone in its cell is a primary requirement for every BTS. Infrastructure in form of a large power supply, a high mounting structure or tower and large high gain antenna(s) for achieving the required coverage are needed at every BTS for the establishment of even a basic level of connectivity.

In disasters situations, infrastructure is prone to failures either by direct destruction or indirectly by factors such as failure of power supply. Additionally, because of these infrastructure requirements, it is infeasible to deploy such communication solutions rapidly.

LifeNet exploits multihop communication to provide coverage over comparable areas with minimal infrastructure. Every device functions both as a host and as a router. Two devices close to each other communicate with each other directly, whereas communication between two far off devices can be relayed by low-power intermediate nodes in a multihop fashion. Infrastructure such as large mounting structures and power suppplies are not required. This facilitates rapid on field roll-out.

© Copyright 2011 LifeNet

No comments:

Post a Comment