El 06/03/2010 19:44, iv3nwv escribió: > Jose, > if you are referring to me I'm not saying that theoretically it is correct to > use as much bandwidth as possible. This is a conclusion you have drawn on > your own. > > Using a 100 kHz bandwith to communicate information at a rate of 1 bit/s > could by sure approach any channel capacity, but the spectral efficiency of > such a communication channel would be quite questionable. Let this option to > NASA deep space communications. > What we need are modes which are both power AND bandwidth efficient. > > I think that the term "spread spectrum" here is misleading. > What's the difference between a communication system which uses a FEC code > with a very low rate, say R=0.01 (one information bit per one hundreds > symbols), and a communication system which hops or spreads the modulating > signal on an equivalent bandwidth? > In my opinion: NONE. > Both systems are using a bandwidth which is one hundreds time the bandwidth > which would be used by an uncoded system. > > The problem is not whether a system is spread spectrum or not. > The problem is how much it is bandwidth efficient. > > Everyone knows that an ortoghonal signalling system approaches the (AWGN) > channel capacity. The legitimate question is if the whole 20 m band should be > used to achieve such a result to communicate information at 3 bit/s. > > For what I know ROS has a really poor bandwidth efficience nor it copes with > MUI (multiuser interference) issues. > I do not doubt that it can achieve an exciting performance under the power > efficiency point of view, but that's not all. > We are called to develop systems which are efficient also in respect to > bandwidth. > > The spread spectrum story is just a bad motivation used against true concerns. > > 73s > Nico, IV3NWV Nico,
Excuse me if I misunderstood it. I believe it is theoretically correct, but not always practical nor possible. For one, I agree that it is incorrect to run over a whole crowded band like 20 meters. You have a point too nobody had made me to stop and think about. FEC or UWB in whatever way, carried to the extremes, are two sides of the same coin. On crowded spectrum, efficiency certainly counts. Nevertheless, it is a complex issue, because I also believe that unprotected systems, like packet has traditionally been is also a waste of bandwidth when a single lost bit sends, say, 255 bytes to trash. As usual, the solution may hardly be on the extremes. 73, Jose, CO2JA
