Problem to be considered
You are asked to design (i.e. choose with justification the detailed design parameters for) a modem for a UHF radio system for use by a search and rescue service. The system is to be used to communicate between a hospital and an ambulance which can travel at speeds up to 150 km/h. Three types of data are to be communicated: (a) digitised ‘telephone quality’ voice, (b) data from life-support instruments and (c) still pictures taken with a digital camera. The system should be capable of operating in both directions will use a handshaking protocol to enable the system to adapt to changing conditions in order to maximise data throughput. Each radio modem will operate on a frequency of around 5 GHz, a bandwidth of 10 kHz and with signal to noise ratios (measured within a 10 kHz bandwidth) exceeding 50 dB for about 20% of the time and exceeding 20 dB for 50% of the time. Fading and increased distances at the limits of the service area will, of course, mean that the SNR often will be considerably lower than this figure and appropriate methods of coping with the degradation in SNR will need to be included in your design study. There is no need to share bandwidth with other communication systems. There is also no need to consider propagation loss models.
The life-support systems produce data at a fixed rate of 400 bits per second, and it is essential that this communication is maintained at all times. In order to minimise the overall data transmission requirements, you should include codecs for both the voice and pictures. Use the library (including electronic publications) and the internet to investigate the state-of-the-art in codecs for these two applications.
How to do the exercise
In making decisions about various parts of the system, you should consider a range (i.e. more than one) possible solutions and you must justify why your final choice it best. It is not sufficient to find the first possibility and to simply use that.
You must support your decisions with appropriate calculations (e.g. to show how all the data fits into the allocated bandwidth) and, if you think it appropriate, support your design choices with appropriate simulations using the MATLAB block-sets. However, it is not necessary to construct the design in Matlab.
The focus of the exercise is concerned with the initial design stages of an engineering system (in this case a modem). Students should work to develop the main design features of the system and it is important to note that there is not a single ‘right answer’. An important aspect of the problem is that you are aware of the advantages and disadvantages of the technology that you select to use, and understand the consequences of your choices. You are expected to engage with the complex situation presented to you and decide what information you need to find and topics you need to learn about in order to develop the design effectively. As a starting point, you may consider the exercises you have done so far in the Digital Communications module. These are all relevant to the problem in hand, but you should also consider what else is important in the design of a digital communications system. Some more complicated features may be irrelevant or too
costly to incorporate into this particular system, in which case you should say so, to show that you have considered them.
It is expected that you will discuss the problems with your colleagues, but the report should reflect your own understanding of the problems and the work is expected to be your own. The submitted work will be checked against other sources online to detect plagiarism.
It may be useful to consider a typical use scenario for the modem. This would be a short statement, outlining the use of the modem over the period of an emergency call-out and saying how much voice and image data would be exchanged. This may help you estimate the requirements. If included, this should go towards the beginning of the report with state diagram.
As a suggested preliminary exercise, you might get together in groups and have a brainstorming session. This requires you to temporarily suspend your critical faculties and just say what comes to mind. Write down all the ideas, even if they sound silly, and then afterwards, go through each idea and fill n details of how it will work, alter it, accept it or reject it, according to the problems that it provokes. The benefit of brainstorming is that it gets you started, and may sometimes generate ideas that you wouldn’t have thought of otherwise. Even if they are not suitable to begin with, they may eventually lead to original solutions. Eventually, each group member will go away with a list of ideas that they then develop into their own solution.
Alternatively, you may want to try an iterative scheme, starting with a familiar idea, such as using BPSK, and go through the questions until you find a problem, such as: ‘The data rate of BPSK is insufficient’. This then provokes another question, such as ‘What modulation scheme will provide a better data rate?’. Then start again with the new modulation scheme until you find a problem with it. You should expect to go through this process several times.
In undertaking the exercise, you should use any information from written sources that you consider appropriate. A list of the papers, websites and books used should be included in your report in a reference list. There are marks for doing this correctly. A common mistake with this exercise is that students write a report detailing, for example, how a Hamming code works or how CDMA works. This detail should be confined to references or an appendix, if necessary. What should go in the report is ‘What are the relevant features of the component you have chosen?’. It is important to note that marks will be given for justified decisions, rather than for background theory.
The final formal report should be understandable by a graduate in electronic engineering and it is suggested that it includes separate sections for each aspect of the system and a ‘conclusions’ paragraph should be added at the end to summarise what has been achieved, what remains a problem (perfection is not expected) and what are the important features of the design. A block diagram is a useful way of showing how all the subsystems fit together. There is no need to repeat any of the information included above. Any figures in the report should, of course be numbered, captioned and referred to in the text and a ‘References’ section at the end should support any . ’Finally, an abstract should be written for the report. This is not an introduction, but a synopsis of your design, written in the fewest words possible. The report should not be more than 15 pages long, but marks will be awarded for content, rather than length.
As a first stage to undertaking the exercise, you should think about possible solutions the following questions.
- What is the maximum error-free data rate that could theoretically be sent through this system?
- What quality is required for voice and for pictures?
- Will the voice channel be full-duplex or half-duplex?
- What error rate would be acceptable for each of the data types to be sent?
- What data rate does each data type require?
- What are the latency requirements for each of the data types?
- Will the different data types have different priorities?
- What will happen to data that is not successfully received for each type?
- What is the reason for using codecs?
- Can the data be compressed?
- What do you think are the relevant parameters for a voice codec? Are these different to an image codec?
- How will these three data types with different volumes and priorities be sent without interfering with one another? What is the best way to combine or multiplex the data types?
- Is it a good idea to put data into blocks or frames, and if so, what size are these?
- What size guard bands should be used?
- Will these priorities change, and if so, will your multiplexing scheme change with them?
- What bandwidth is required ideally for each of the data types to be sent?
- Of the several error correcting codes investigated, what are their relative merits?
- What error correction coding (ie. channel coding) will you use?
- What processing should be carried out on each data type separately, and what can be done to the multiplexed data as a whole?
- What modulation scheme is to be used? Can a single modulation type be employed in such a way as to maximise data throughput in all situations?
- If the modulation type changes, what measurements would provoke this change, and how will both ends of the link be aware of the change? What handshaking procedures take place?
- What minimum and maximum speeds would you specify for the vehicle while in use on emergencies?
- In the case where the vehicle is stationary in a dead-spot (i.e a location which has poor reception), what diversity technique might be used to improve signalling?
- What is an appropriate interleaver length? Is this the same for all types of data sent?
- Will the receivers be coherent or non-coherent?