Switching problem

[lokmanhi]

Hi.
I have one question regarding to radio communication. We are installing digital radio link ( for data & voice comm.) between an oil processing ship (Floating Production,Storage & Offloading - FPSO) and remote platform at offshore. The FPSO stationed at one place but will move in various directions - rotating the turret, up and down, etc. The radio installed is 400Mhz UHF radio 1+1 hot standby, which will switchover in order to obtain the best signal. Now, it is discovered that the switching will cause data loss from the remote platform. Futher more, due to ship's movement, it is afraid that the switching will occur at all time, thus cause the any data cannot be sent to the ship. Do you have any suggestion to resolve the problem?
Thank You.

 

[robertjo24]

In the past, the only reason I knew of to have a hot standby for 400 Mhz was in case the primary system went off line. The transmitters would be set up in hot standby mode to switch on failure, with the receivers both listening at the same time. There would be a receiver combiner that would select which signal to use according to quality (noise content). This is a fairly common set up for a 400 Mhz system.
Are you running completely separate redundant systems on separate antennas? Are you utilizing space diversity for this signal? I am having a hard time understanding why the system would undergo several switches.
Any more info you can provide?

 

[lokmanhi]

These are description:

1) The radio link is designed for a ship that stationed at offshore at one fix point called mooring point. She is allowed to rotate 360 degree all time depends on whether condition. The link is from the ship to two remote platforms at distance of 2 km & 16 km each.

2) The system is designed as ' hitless switch' condition which no interruption shall occur during switching between the two radios

3) The redundancy is two separate units with 2 different antennas. This designed is to cater the movement of the ship; pitching, rolling, rotating at all axis.The transceivers are to obtained the best signal so that there would no data loss at all during the ship's movement.

4) Yes, it is designed with the space diversity.

5) The problem that I faced is there was switching delay for about 5 sec and more during transmission when the equipment tested onshore, not yet offshore. Offshore condition might worse, which can always cause loss of signal in all time, as explained before due to inconsistent of ships position.

My futher question:
1) Is 400 Mhz suitable for this kind of situation or any other suitable frequency to cater the problem?
2) Until now I cannot figured out what is the main problem, is it normal that the switching will cause slight delay or because of components; such as mux, etc

Thanks.

 

[BrianR]

I think you will get signal loss no matter what the frequency. Multi-path reception will be the main cause and I think you will need two complete parallel systems in operation to minimise data loss.

 

[lokmanhi]

Yes, the system is two parallell system. But is it possible to get the system as per design, hitless switching (no interruption of the signal transmission and receiving) based on the ships condition? Since our major concern is the signal loss,which can cause loss of oil production from the remote platforms and also at the oil processing ship.

 

[ragudelo]

Believe there is no satisfactory solution for your problem, the way the system was designed.
First, you must understand that a hitless system with a moving target is simply impossible (actually, it's two, since the platform will move significantly too). No matter the current technology used, there will be either packet or frame losses. By using modern communication techniques, it's possible to minimize the impact, but will always be there.
Now, regarding your application, if your control system is not able to endure the failures of the communication link, then the system design is wrong. You must know that, it's a fairly common system design rule in the oil industry.
Second, you must understand that no matter the antenna design at the moving target, there is no physical way to guarantee that you will have stable signal levels in any position of the target. A fast servo may help a bit, but will be slower than nature (the mass of your antennas at 400 MHz will be a high load to move fast). Of course, you may try with smart antennas, will be an interesting experiment (but not impossible, just VERY expensive; but oilers may afford such gadgets).
Finally, gee A 400 MHz digital radio? Which one? Any way, no matter what you try to install there; with a moving target, the signal changes will be continuous and would keep the system switching endlessly.
Suggestion: Get back to the design table, and make sure the control system is smart enough locally to keep control of the production, no matter what happens to the link. Then you can use remote communications to run heavier optimizing algorithms and tune the local control system. For voice, try using VoIP, it might be more resilient than standard TDM lines, providing enough bandwidth is allocated to them.
RA

 

[BrianR]

I am not experienced in oil platform work and am not familiar with the licencing requirements for the radios, however it sounds like you are using directional antennas on a servo platform. If your comms licence permits it, for 16Km range then how about using higher powered transmitters with low gain, omnidirectional, vertically polarised antennas on fixed mounts and situated as high as possible on the vessel. They will suffer degradation with cross polarisation as the vessels roll but even cross polarisation up to 45 degrees will not kill the signal and so you will not need a levelling servo.

Any switching between systems must be done electronically (read microseconds speed) at the data level with an error correcting protocol to resend lost packets. There will always be momentary loss of data packets from both systems even with space diversity antennas. If this is intolerable then add a third system, probably in the VHF band as well as the UHF. VHF will suffer a completely different pattern of fading under the conditions you mention.

Of course sattelite works well too.

 

[VEBill]

Interesting problem. This is the same sort of problem that has already been solved many times before.

From the RF point of view, you would need an active antenna combiner. An example is the ViaSat QDC-100. Be advised that this is a very expensive unit, but the point is to show that a solution is technically possible at the RF side if you really need it. Of course, you'll probably still have occasional RF outages even with such a combiner.

http://www.google.com/search?q=QDC-100

The other option (much cheaper) is to address the data side. By using a packet radio controller, then the short radio outages can be dealt with. The packet radio controller will simply resend the missing data packets. This assumes that the increased latency is acceptable.

Example:

http://www.google.com/search?q=VDC-500

There are other similar products from other companies.

If the RF outage is too long, then you'll get an error message and your application SW should deal with it.

 

[lokmanhi]

Thanks a lot for your suggestion

 

[sierraromeo]

The switching unit for the RF has a built in combiner thus will have a single antenna to radiate the signal. The radio operate as 1+1 with QAM transmission with a subrate multiplexed TDM data input to both the radio.

The multiplexer used acts as the master clock for both the radios. When configured as 1+1, only one of the radio is "hot" while the other is not. When the switching unit operates the mux had to re-synchronize to the operating radio at the local end, during this re-synchronizing period the input signal is lost.

The remote end, clocking is set to follow the received signal to ensure correct data detection which is normal for any synchronous data transmission. The radio switches witout any problems but the mux needed time to reestablish it's synchronization.

With that in mind, wether we use a 1+1, frequency diversity or a space diversity system as long as the mux had to re-synchronize the down time is unavoidable.

I feel that it the solution lies with a mux that has a duplicated 2Mbit/s output connected to two separate radio running at 1+0 configuration with different frequencies. The mux then decides which signal is the most stable. This way we avoid any re-syncronization during radio failure.

With the syncronization down time out of the way, the next thing is to consider the reflection effect on the signal. A pathloss calculation can provide answer on the power budget but it does not give you the reflection effect on the signal.

There are other issues that need considering but let's tackle this one first.

 

[fgiordiano]

There are some options other than switching to reduce lost data.
First, you can increase the power with a linear amplifier and use an omnidirectional antenna. This will eliminate the need for beam tracking.
Also applying some forward error correction, like Viterbi coding or others, can greatly reduce the signal strength needed in order to avoid lost packets. There are many modem manufacturers that can supply such equipment (Hughes Network Services, L3 communications, and so on).
Power and coding can overcome almost any path loss issue. Reflections and multipath effects can be mitigated with antenna diversity.