 Let's extend our understanding about the G-pons. First, we'd look at the principle and then we'd look at the specifications in terms of distance which are recommended for G-pons from the traffic type and from the physical limitations of the optical cable in terms of attenuation and absorption. And then we'd finally look at the architecture. First and foremost, G-pons has something different from its predecessors that is it actually adapts wavelength division multiplexing that is it chooses different wavelengths for the upstream and the downstream. Each wavelength can carry TDMA time division multiplexed data from various users. So when it comes to multiplexing, then G-pons utilizes either of the two mechanisms. Number one, as far as the downstream traffic is concerned, the traffic is sent as broadcast. It means that the passive splitter is going to simply repeat or replicate the traffic which is coming from the central office. As far as the upstream is concerned, time division multiplexing or time division multiple access mechanism is utilized from the user equipment to the central office. The assignment of these time slots is a different subject and it is left to the higher layers or the network manager to adopt a certain type of TDMA mechanism. A G-pon makes certain recommendations in terms of the distance. For instance, although in terms of logical understanding, if the signal affords up to 60 kilometers coverage can be used between the optical network unit or the ONT on the user side and the optical line transmission unit on the central office or the operator side. But this 60 kilometers distance is only logical because the maximum physical constraints are determined by the signal degradation attenuation and also by an interesting phenomenon. That is, if you want to adopt this G-pon on the access side as a technology to offer multiple user services like voice, video, data and interactive applications, then the delay on the signal transfer has to be less than 1.5 milliseconds. That is the propagation delay. A propagation delay that increases this particular bound then starts to have an observable impact on the other end. So logically it can be extended up to 60 kilometers but physically it has limitations. Those limitations also include the traffic type that is being carried between the two endpoints. We'd look at the G-pon architecture in a little more detail with some variants. Let's first of all look at some of the abbreviations which have been used. We have well known fiber to the home, fiber to the cabinet, fiber to the curb. These are quite well known but an interesting observation is the choice of combination which is made between the OLT and the ONU. As you can see here for the sake of clarity the OLT is shown as a single module but the ONU or the ONT is shown as comprising two parts the network unit and the termination unit. So the network unit is what deals with the network related functionality and the termination module is the one that deals with interfacing or the traffic termination issues. It is simply what we can say is there is one aspect which is related to the dropping of signal more related to the physical and data link layer endpoints but when it comes to the network unit it deals with the complexities which arise because of the fact that each user is connected to a network that OLT is trying to manage at the same time. It is essentially a many to one transmission where we have one OLT and many ONUs and ONTs. So essentially this makes the choice dependent upon the vendor and the particular kind of application. Here we see we have on the right most side we have a building we have individual subscribers so depending upon the user requirements of how the user wants to be serviced correspondingly different kinds of combinations can be realized.