The simple schema above illustrate the (very) basic concepts underlying communication technology. We will describe it in very simplistic terms to understand the challenges that communication systems face when transmission is performed over a physical medium.
In a nutshell, an information source producing specific messages wishes to transmit them to a remote destination without loosing the intended meaning. The transmission must be performed over what is commonly called a channel. The channel can be viewed as a transport medium augmented with the process of encoding the produced messages into transmittable signals and decoding these signals back into meaningful messages. The objective of the system is to optimize the encoding depending on the information source without diminishing its expressiveness, reliably transmit signals over the channel from the encoder to the receiver without loosing too much information and accurately decode the signals back into a representation which is meaningful to the destination source. Analysis of the information source can (must) be performed prior the implementation of the communication system. Information sources can be divided into two types : discrete source (ex.text) and continuous sources (ex.voice). Without going into too much details, the concept of entropy arises from the analysis of information sources and serves as a measure of information and information rate. The entropy is the link between the abstract representation of a message and the physical limits that a channel must conform to in order to efficiently transmit information. The capacity of a channel is then measured by the amount of information it transmits (produced by the information source) and is therefore not reduced to the only number of symbols it can transmit. A physical transport medium introduces interferences such as noise that can distort the message traveling between the sender and the recipient. A noisy channel must take into account the characteristics of its noise source in the measure of its capacity.
This is a very quick overview of information theory. Please refer to - A Mathematical Theory of Communication by Claude E. Shannon- for more in-depth information on the subject.
As an example of a communication system implementing the basic concepts presented in the previous section we propose to have a look at the infrastructure of the Internet. The figure above is a layered schema of the systems involved in a computer network.
Since we are concerned about networking transports, we are mostly interested in the bottom three layers : Physical, Data Link and Network.
The Physical layer is responsible for encoding bit streams into signals, transmitting these signals over a physical medium (wire, wireless) and decoding these signals back into a binary format. It encapsulates the channel concept previously described earlier. The physical layer provides interfaces to electromagnetic mechanisms by which the transmission of binary data is performed over a specific transmission medium. The type of modulation to be used to encode data, the frequency to broadcast from and the coding techniques are all specifications which are related to the physical layer.
The Data Link layer is accountable for the transfer of data between network entities (hop to hop) and mechanisms to support multiple access to the network channels. Some implementation of the Data Link layer also provide error detection and correction. More importantly the concept of address and data frames are introduced by this layer. The Data Link layer frames the data to be transmitted over the channel with an header and a footer which contains information about the source and destination addresses (MAC addresses). Other important aspects of the Data Link layer are the mechanisms in which it monitors access to the network channel. Generally multiple computer systems share the same channel over which transmission of data is performed. The Data Link layer is responsible of determining who is allowed to access the channel at a any given time. This is called channel sharing and multiple techniques are possible. To name a few : token-ring, master-slave, CSMA/CD are usually used for wired channels and CSMA/CA, slotted ALOHA, CDMA, OFDMA for wireless channels. As an example Ethernet uses CSMA/CD.
Finally the Network layer is responsible for the end to end delivery of packets in the network. This layer is more robust in terms of addressing (IP) and delivery then the Data Link layer. Of course the Network layer is built on top of the Data Link but whereas the Data Link layer was concerned with single hop transfers, the Network Layer is specifically designed to route and forward packets over the whole network.
It is important to understand that the internet was not build around this specific layered representation of networking. It is simply convenient way of describing different system interactions that may be used to implement a specific functionality but not mandatory.
 reflection
|
 |
Internet 0 is a powerful implementation of a communication system intended for small embedded devices to be deployed within the existing Internet infrastructure. As we will see in the next paragraphs, it cleverly merges the three layers previously described from the OSI model and defines a new modulation scheme in which the representation of data is carried from end-to-end.
Each nodes present in an Internet 0 network have the ability of generating their own IP address and broadcast it to the whole network using UDP packets. The nodes interact in a peer based manner and therefore do not need intermediate servers in order to communicate. I0 is based on the principle that the network defines itself from its leaf nodes. This simple concept is known as the end-to-end principle.
One of the most important aspects of the I0 system is its modulation scheme. Whereas the speed and distance of data transmissions is important in a computer network an IO embedded device network is less concerned about it. This simple observation has a huge impact when considering a modulation scheme and leads to the concept of small bits versus big bits networks. In a computer network electromagnetic signals travel at a speed comparable to the speed of light (300 m/us). Each bits transmitted through the channel are smaller than the network itself (small bits) and therefore need to be modulated with a carrier. In such small bits networks impedance matching needs to be performed at each ends of the channel in order to eliminate signal reflection (portrayed in the figure above). In the case of I0 these limitations don't hold anymore. Usually the network will be smaller than the bits it transports (big bits) leaving time for a bit to settle over the transmission medium. This settling period is commonly called the transient response (portrayed in the above image) and its duration may change from medium to medium. Therefore I0 bits are encoded into series of impulse response (with appropriate width depending on the transient response of the transmission medium) and decoded as bits when a change of state of the transmission line is detected - similar to serial communication. All the modulation is performed in the time domain (bit = event of an impulse) and is carried through out the network.
Considering that I0 leaf nodes are responsible for their own IP addressing and transmission of UDP packets coupled with the fact that modulation is performed from node to node - data and modulation is always kept intact from end-to-end in an I0 network. Furthermore layers 1 to 3 in the OSI model are merged together. An I0 node is a modem and a router capable of handling signals and protocols.