4. Monitoring System Upgrade

(1)

4. Monitoring System Upgrade

Here we describe all the interventions that led to the upgrade of the original Monitoring System.

(2)

4.1 – Why making an upgrade on Rambouillet_1?

At the start of this project, the monitoring system had few problems; let’s begin to explain the ones on Rambouillet_1: remember that there were 6 Monitoring Devices for getting measures over carriers received by 12 antennas.

The first problem is an operational one: since during a lineup session, at least 2 Spectrum Analyzers are used (one for Co-Pol and one for Cross-Pol components), good resolution is assured if there are till 5 measures on different carriers simultaneously; over this quantity, an operator can’t distinguish the variation of the monitored spectrum without waiting lots of seconds between two consecutive screenshots.

In addiction, every time a lineup session is started, the background system is suspended, and this can affect the time percentage during which it works. The Image

4.1 shows a typical usage of a Spectrum Analyzer in the old system; we can observe

that most of the time it was busy for lineup work. We were able to find similar values also for the other devices.

(3)

(Image 4.1 – Typical statistics for SA before upgrading)

The main consequence of it was that the time between two consecutive background measures of the same carrier was neither constant nor near the 10 minutes the optimal system can perform. As it is a very important index for evaluating system performances, in the next paragraph we’ll show it before the upgrading, in order to be able to make a comparison after having added the new hardware.

Another important aspect to consider is that the original system wasn’t scalable at all: actually, adding a new monitoring antenna was impossible because any input of RF

(4)

Finally, there was the necessity to remove old switches that used the GPIB standard for the communication with the other devices; since the fact that the mutual bus for the interface with the Front End Controllers was on a common Ethernet LAN, a GPIB/LAN Gateway was necessary for each of them; actually, the new ETL™ hardware installed is directly controlled via TCP/IP protocol through an Ethernet port.

4.2 – Original statistics for Rambouillet_1

In order to evaluate performances of the Background System, we’ll consider 3 reference carriers. For each of them, we’ll use the time between two consecutive measures as a sensitive marker and we’ll show its Probability Density Function,

Cumulative Density Function, and other common statistics.

For evaluating performances before upgrading, we considered a period of 7 days, from September 21st at 14:22:00 to September 28th at 13:00:00. This choice isn’t casual, because we selected a period in which any particular event that could lead to wrong conclusions occurred, like, for example, exceptional news or sport events that required too much lineup sessions.

(5)

Carrier # 1 Carrier # 2 Carrier # 3

ID NOORSAT M6 CETEL Satellite AB2 AB3 EB4 Transponder B1 KA01 B1 Carrier Type Digital TV Analog TV Digital TV

Uplink CF: 13773.67 [MHz] 13768.00 [MHz] 14256.59 [MHz] Occupied Bandwidth 36000.00 [KHz] 36000.00 [KHz] 1843.20 [KHz]

Nominal EIRP 49.27 [dBW] 53.64 [dBW] 30.89 [dBW] C/N0 103.24 [dBHz] 101.48 [dBHz] 79.57 [dBHz]

(Table 4.1 – List of sample carriers)

Remember that all the data used in this chapter are collected via the SIECAMS® interface, the database of which contains stored values till 1 year in the past.

4.2.1 – Original data for Carrier #1

Here below statistics on the first carrier are listed. The same formulation is used for the others.

(6)

Index Value for Carrier #1

Total Number of Background measures in the selected period (One week) 313 Minimum value of minutes between two consecutive measures 12 [Min] Maximum value of minutes between two consecutive measures 94 [Min] Average value of minutes between two consecutive measures 27.22 [Min] Standard deviation value of minutes between two consecutive measures 12.57 Percentage of two consecutive measures within 10 minutes 0 [%] Percentage of two consecutive measures within 15 minutes 14.70 [%] Percentage of two consecutive measures within 20 minutes 22.36 [%]

(Table 4.2 – Statistics for carrier #1)

(7)

(Image 4.3 – CDF for carrier #1)

4.2.2 – Original data for Carrier #2

(8)

Total Number of Background measures in the selected period (One week) 310 Minimum value of minutes between two consecutive measures 12 [Min] Maximum value of minutes between two consecutive measures 92 [Min] Average value of minutes between two consecutive measures 27.47 [Min] Standard deviation value of minutes between two consecutive measures 12.03 Percentage of two consecutive measures within 10 minutes 0 [%] Percentage of two consecutive measures within 15 minutes 12.26 [%] Percentage of two consecutive misures within 20 minutes 20.32 [%]

(9)

4.2.3 – Original data for Carrier #3

(10)

Total Number of Background measures in the selected period (One week) 331 Minimum value of minutes between two consecutive measures 06 [Min] Maximum value of minutes between two consecutive measures 102 [Min] Average value of minutes between two consecutive measures 27.49 [Min] Standard deviation value of minutes between two consecutive measures 12.98 Percentage of two consecutive measures within 10 minutes 0.3 [%] Percentage of two consecutive measures within 15 minutes 12.99 [%] Percentage of two consecutive misures within 20 minutes 22.66 [%]

(11)

4.3 – Why making an upgrade on Rambouillet_2?

As far as the second sub-site, the main problem was that two different systems coexisted together, making its administration harder. Actually, having them meant configuring different paths for antennas on the management interface, doubling path calibration and all problems of any nature for the same hardware.

Furthermore, the original configuration didn’t permit to add efficiently new resources, like antennas for monitoring new satellites, because this action could make performances worse. Having 3 spectrum analyzers to monitor all antennas was