How to Build a DMR Simulcast Network

How to Build a DMR Simulcast Network Workshop: Digital Mobile Radio Association DMR Alessandro Guido DMR System Design, Leonardo S.p.A. Augusto Colombo DMR Sales Technical Support, Leonardo S.p.A. 1

Broadcast: Simulcast vs Cellular F1, F2 F7, F8 F7, F8 F1, F2 F1, F2 F1, F2 F3, F4 SIMULCAST: same channel used by neighbouring RBSs, only 2 frequencies for the whole system F1, F2 F3, F4 F5, F6 F7, F8 CELLULAR: different channels used by neighbouring RBSs (cluster) 2

Simulcast Key Factors Issue: to manage radio communications in overlapping areas Solution: implementation of effective procedures for synchronisation, equalisation and voting Result: the radio network acts as a single virtual repeater synchronisation equalisation voting Virtual repeater MAIN BENEFITS: The coverage area is actually improved, compared to the repeaters single contributions. Only one frequency pair RF spectral efficiency cost savings for licensed frequencies Optimal coverage of wide areas Easy-to-use Same operations like under a single site repeater 3

Simulcast Key Points Single Virtual Repeater Wide area coverage Requires a low number of frequencies in the network Intrinsic call Hand-Over Network Coverage easy to be increased 4

General Hypotheses for a Radio Coverage Study When evaluating the radio coverage for a simulcast network, that uses multiple synchronized and interconnected repeaters, it is essential to know the following parameters: exact location of repeater sites (lat., long., height from sea level) antenna height from ground ( i.e. tower height) accuracy and resolution of the terrain, building, and clutter data propagation model used choice of time/location/margin values physical characteristics of the study area (plains, mountains, urban, etc.) allocated frequency bands radio electrical characteristics of the radio equipment used ( both fixed and terminal) radio electrical characteristics of the antenna equipment used ( both fixed and terminal) requested service: portable and/or mobile, outdoor and/or indoor, in-car and/or mobile Only by knowing all the above parameters it is possible to simulate with a good approximation the service area (radio coverage). The service area is where a level of radio signal able to ensure the requested service is guaranteed. Since the radio propagation can be described by a statistical process, this requirement is satisfied only with a certain probability, called location probability, where the signal level is enough to ensure the requested service in the service area REQUIREMENTS: a sufficient value of the local average RF power a given coverage probability 5

DMR Simulcast - Radio Coverage project guidelines standards These standards are intended to provide information that allows accurate and consistent modeling of coverage performance. These standards utilize specified and measured product performance and other sources that provide a sound engineering basis to the values suggested. Calculation hypotheses 1. Calculation of the average path loss in quasi-smooth terrain, no obstructions 2. Diffraction losses (orography), roughness of the terrain => database 3. Local territorial morphology, Land Use => database models and mathematics algorithm for radio coverage 4. Power Budget, evaluation of useful signal for communication; 5. Additional losses to reach the coverage probability for the requested service. Simulcast Delay Spread Simulation vs real coverage verification radio equipement radio coverage probability Hess method 6

Calculation Hypotheses - steps 1, 2 and 3 1. Calculation of the average path loss in quasi-smooth terrain (Okumura Hata/Davidson Ext.) TIA recommends in TSB-88 a modification of the Hata model to cover a wider range of input parameters. The output parameter of Olumura Hata/Davidson Extended algorithm is the average path loss LdB calculated according to the terrain type. In addition to this extension, this model can optionally include additional path loss due to reflections and diffraction loss from terrain obstacles. 2. Diffraction losses (orography) (Epstein Peterson Diffraction) Diffraction attenuation is due to orographic obstacles due to local territory morphologies. A rich and dedicated Digital Terrain Model (DTM) database can be used together with the Epstein and Peterson algorithm. Recommended resolution not less than 1 arcsec (~30 meters). 3. Added attenuation due to local territory morphologies Clutter Added attenuation due to local territory morphologies to take into account urban and suburban density, the type of herbaceous or woody vegetation, rocky areas, glaciers, etc. There are different databases that can be used: - Land Use and Land Cover (LULC) database from the United States Geological Survey (USGS) - CORINE (COoRdination of INformation on the Environment) Land Cover - European Urban Atlas: it provides reliable high-resolution land use maps for 305 Large Urban Zones (i.e. with more than 100.000 inhabitants) and their surroundings. They incorporate different data formats and at least 15 classes; resolutions are up to 1 arcsec (~30 meters). 7

Calculation Hypotheses - step 3 - terrain database exmples Terrain Orography The Digital Elevation Model normally used has a resolution plan (XY) of about 50x50 meters, and a height resolution (Z) of about 10 meters, and is built specifically for the coverage prediction tool ROME 8

Calculation Hypotheses - step 3 - terrain database exmples Terrain Morphology The Clutter Land Use normally used reproduces the morphological classes of territory, with a resolution plan (XY) of about 50x50, meters Different colors are associated to different terrain use: Urban Atlas database ROME 9

Calculation Hypotheses - step 3 - terrain database exmples Terrain Morphology Land Use Land Cover (LULC) from United States Geological Survey (USGS) BOSTON 10

Calculation Hypotheses - steps 4 and 5 4. Link Budget calculation Link budget is the calculation of the local average power, usually corresponding to a 50% coverage probability, obtained by summing gains and losses related to TX RBS, RX RBS and terminals. Link budget calculation assures the correct balance between uplink (Talk-In) and downlink (Talk-Out) paths; its output is the maximum allowed attenuation. The worst case, i.e. the one with minimum path loss that still guarantee the communication is used to identify the coverage area, i.e. the service area where terminals can actually send and receive with high reliability and good quality. 4.1) Delivered Audio Quality DAQ (TIA TSB-88) These documents contain recommendations for both public safety and non-public safety performance that are intended be used in the modeling and simulation of these systems. These documents also satisfy the need for a standardized empirical measurement methodology that is useful for routine proof-ofperformance and acceptance testing and in dispute resolution of interference cases that might emerge. 5) Coverage reliability calculation (90 or 95%) Coverage reliability is assumed as a function of radio service. The recommended values are as follows: 11

Link Budget Calculation - 1 A mobile radio channel experiences short-term variations due to multipath fading. This has undesirable effects on the audio quality or BER of a wireless link. Propagation models of practical use do not predict this variation. Its impact is included within the sensitivity of the receiver via a parameter called the faded Cf/(I+N) ratio. This ratio depends on the modulation type, coding and error correction details and the rate of fading. The analysis process requires the following parameters: 1. Noise floor 2. Static carrier-to-noise ratio (Cs/N) 3. Required Cf/(I+N) (faded performance threshold) 4. Reliability margin In order to determine the noise floor of the receiver, subtract the static carrier-to-noise ratio (Cs/N) from the reference sensitivity (12 db SINAD Analog, 5% BER DMR Digital): reference sensitivity (DMR = -118dBm typical) _ Modulation Type Analog FM (12,5kHz) Cs/N (db) Source 7 db TSB-88 DMR 5.3dB TSB-88 12

Link Budget Calculation - 2 Determining the required faded carrier to interference plus noise ratio Cf/(I+N) at a given Delivered Audio Quality (DAQ): Modulation Type Analog FM (12,5kHz) DAQ 3.0 DAQ 3.4 DAQ 4.0 Source 23.0 db 26.0 db 33.0 db TSB-88 DMR 14.3dB 15.6dB 19.4dB TSB-88 Determining the faded performance threshold (e.g. for DAQ 3.4): Typical body loss (from ITU-R P.1406): -107,7dBm -118dBm Dynamic Threshold (Faded Performance Threshold) Static Threshold (Reference Sensitivity) Noise Floor Cf/(I+N) = 15,6dB Cs/N = 5,3dB 13

OL-02-004 OL-02-010 JE-02-004 PV-02-008 SU-02-001 SU-02-002 SU-02-003 OL-02-004 OL-02-010 JE-02-004 PV-02-008 SU-02-001 SU-02-002 SU-02-003 Link Budget Calculation - 3 Body loss and inc-ar loss down link Required Delivered Audio Quality Reference Sensitivity Faded Sensitivity Power Budget Talk OUT 1 2 3 4 5 6 7 Power Budget Talk IN 1 2 3 4 5 4 5 Channels Type Česká republika Česká republika Česká republika Česká republika Česká republika Česká republika Česká republika Range VHF VHF VHF VHF VHF VHF VHF Antenna TX RX #1 #1 #1 #1 #1 #1 #1 Antenna Type Omnidirez. Omnidirez. Omnidirez. Omnidirez. Omnidirez. Omnidirez. Omnidirez. Model RF331NH RF331NH RF331NH RF331NH RF331NH RF331NH RF331NH Antenna Azimut N - - - - - - - Antenna Cable Type Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Model LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN Antenna Height m 53 30 35 8 30 20 30 Antenna Cable Length m 83 60 65 38 60 50 60 DMR TX Power W 25,0 25,0 25,0 25,0 25,0 25,0 25,0 DMR TX Power dbmw 44,0 44,0 44,0 44,0 44,0 44,0 44,0 Combiner Loss db 0,00 0,00 0,00 0,00 0,00 0,00 0,00 Circulator Loss db 0,40 0,40 0,40 0,40 0,40 0,40 0,40 Jumper Loss db 0,60 0,60 0,60 0,60 0,60 0,60 0,60 Duplexer Loss db 1,20 1,20 1,20 1,20 1,20 1,20 1,20 TX Power W 15,1 15,1 15,1 15,1 15,1 15,1 15,1 Antenna Gain dbd 4 4 4 4 4 4 4 Cable Loss db 2,38 1,72 1,87 1,09 1,72 1,44 1,72 TOTAL TX Inse. Loss db 4,58 3,92 4,07 3,29 3,92 3,64 3,92 ERP dbmw 43,4 44,1 43,9 44,7 44,1 44,3 44,1 ERP W 21,9 25,5 24,6 29,4 25,5 27,2 25,5 Portable Mobile Talk OUT Loss Path Loss TX Power W 5 5 5 5 5 5 5 Antenna Gain dbd -2-2 -2-2 -2-2 -2 Refer. Sensitivity dbmw -118,0-118,0-118,0-118,0-118,0-118,0-118,0 Faded Sensitivity dbmw -107,7-107,7-107,7-107,7-107,7-107,7-107,7 TX Power W 10 10 10 10 10 10 10 Antenna Gain dbd 0 0 0 0 0 0 0 Refer. Sensitivity dbmw -118,0-118,0-118,0-118,0-118,0-118,0-118,0 Faded Sensitivity dbmw -107,7-107,7-107,7-107,7-107,7-107,7-107,7 Body Loss db 16 16 16 16 16 16 16 In Car Loss db 6 6 6 6 6 6 6 Buinding Loss db 0 0 0 0 0 0 0 TOTAL Mar. Pt. IC db 22 22 22 22 22 22 22 TOTAL Mar. Pt. OD db 16 16 16 16 16 16 16 TOTAL Mar. Mb. db 0 0 0 0 0 0 0 DownLink Pt. IC db 127,1 127,8 127,6 128,4 127,8 128,0 127,8 DownLink Pt. OD db 133,1 133,8 133,6 134,4 133,8 134,0 133,8 DownLink Mb. db 151,1 151,8 151,6 152,4 151,8 152,0 151,8 Portable IN CAR Talk OUT Signal Threshold Portable OUT DOOR Talk OUT Signal Threshold Mobile Talk OUT Signal Threshold dbmw dbmw dbmw -85,7-85,7-85,7-91,7-91,7-91,7-85,7-85,7-91,7-91,7-107,7-107,7-107,7-107,7-107,7-85,7-85,7-91,7-91,7-107,7-107,7 Channels Type Česká republika Česká republika Česká republika Česká republika Česká republika Česká republika Česká republika Range VHF VHF VHF VHF VHF VHF VHF Antenna TX RX #1 #1 #1 #1 #1 #1 #1 Antenna Type Omnidirez. Omnidirez. Omnidirez. Omnidirez. Omnidirez. Omnidirez. Omnidirez. Model RF331NH RF331NH RF331NH RF331NH RF331NH RF331NH RF331NH Antenna Azimut N - - - - - - - Antenna Cable Type Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Cellflex 1/2" Model LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN LCF12-50JFN Antenna Height m 53 30 35 8 30 20 30 Antenna Cable Length m 83 60 65 38 60 50 60 DMR Refer. Sensitivity dbmw -118,0-118,0-118,0-118,0-118,0-118,0-118,0 DAQ 3.4 Faded Sensitivity dbmw -107,7-107,7-107,7-107,7-107,7-107,7-107,7 Combiner Loss db 0,00 0,00 0,00 0,00 0,00 0,00 0,00 Circulator Loss db 0,40 0,40 0,40 0,40 0,40 0,40 0,40 Jumper Loss db 0,60 0,60 0,60 0,60 0,60 0,60 0,60 Duplexer Loss db 1,20 1,20 1,20 1,20 1,20 1,20 1,20 RX Add. Loss db 0,0 0,0 0,0 0,0 0,0 0,0 0,0 Antenna Gain dbd 4 4 4 4 4 4 4 Cable Loss db 2,38 1,72 1,87 1,09 1,72 1,44 1,72 TOTAL RX Inse. Loss db 4,58 3,92 4,07 3,29 3,92 3,64 3,92 Frequency Range VHF Portable Mobile Talk IN Loss Path Loss TX Power W 5 5 5 5 5 5 5 Antenna Gain dbd -2-2 -2-2 -2-2 -2 Refer. Sensitivity dbmw -118,0-118,0-118,0-118,0-118,0-118,0-118,0 Faded Sensitivity dbmw -107,7-107,7-107,7-107,7-107,7-107,7-107,7 TX Power W 10 10 10 10 10 10 10 Antenna Gain dbd 0 0 0 0 0 0 0 Refer. Sensitivity dbmw -118,0-118,0-118,0-118,0-118,0-118,0-118,0 Faded Sensitivity dbmw -107,7-107,7-107,7-107,7-107,7-107,7-107,7 Body Loss db 10 10 10 10 10 10 10 In Car Loss db 6 6 6 6 6 6 6 Buinding Loss db 0 0 0 0 0 0 0 TOTAL Mar. IC db 16 16 16 16 16 16 16 TOTAL Mar. OD db 10 10 10 10 10 10 10 TOTAL Mar. Mb. db 0 0 0 0 0 0 0 UpLink Pt. IC db 126,1 126,8 126,6 127,4 126,8 127,1 126,8 UpLink Pt. OD db 132,1 132,8 132,6 133,4 132,8 133,1 132,8 UpLink Mb. db 147,1 147,8 147,6 148,4 147,8 148,1 147,8 Portable IN CAR Talk IN Signal Threshold Portable OUT DOOR Talk IN Signal Threshold Mobile Talk IN Signal Threshold dbmw -84,7-84,7-84,7 dbmw -103,7-103,7-103,7 Body loss and in-car loss up link dbmw -90,7-90,7-90,7-90,7-84,7-84,7-90,7-103,7-103,7-84,7-84,7-90,7-90,7-103,7-103,7 Talk Out Signal Threshold 28 November Worst case 2017 Talk In Signal Threshold 14

Coverage reliability calculation Coverage reliability is assumed as a function of radio service. The recommended values are as follows: Radio Service Faded Performance Threshold Public Safety DAQ 3.4 LMR DAQ 3.0 The reliability is the probability that the received local average signal strength predicted in a given tile equals or exceeds the desired Faded Performance Threshold. Link budget gives a 50% coverage probability to have a local average power better than the signal threshold. To have a much more accurate coverage probability, it is needed to take into account two others parameters: local average power statistical law: to find the space and time probability in which the signal level is above the required minimum level standard deviation: data dispersion from average value measurement normal distribution as a function of probability (ITU P. 1546) additional attenuation for local territorial morphology probability of coverage 50% + ( standard deviation on the local morphology x ) 15

DMR Simulcast Delay Spread - 1 Two signals transmitted at precisely the same time will arrive at a receiver at different times, depending on the distance that the signal travels. This difference in arrival times of the two signals time is known as delay. The combined effect of multiple signals in a simulcast environment can be computed using the "Delay Spread" factor. A method of quantifying modulation performance in simulcast and multipath environments is desired. Hess designed a model, called the multipath spread model. The model is based on the observation that for signal delays that are small with respect to the symbol time, the bit error rate (BER) observed is a function of RMS value of the time delays of the various signals weighted by their respective power levels. This delay spread represents the entire range of multipath possibilities to a single number. The multipath spread for N signals is given by: Since BER is affected by Tm more strongly than N, any value of N can be more simply represented as if it were due to two rays of equal signal strength. This would be interpreted as Tm can be calculated by evaluating for two signals, where P1 equals P2. The value for Tm is the absolute time difference in the arrival of the two signals. 16

DMR Simulcast Delay Spread - 2 In the simulcast delay spread Hess studies, two parameters are used: the Capture Margin and the Maximum Simulcast Delay Spread value: "Capture Margin": if any signal exceeds at least one from all the others by this value, then this single signal is the only one recognized by the receiver. If the capture margin criterion is not met, the delay spread value is computed with the multipath spread delay model "Maximum Simulcast Delay Spread (SDS)": it is the maximum permitted delay value. If the computed delay spread exceeds the maximum value, then the point is considered to have no usable signal. If the computed delay is less than the maximum specified delay, the signal at the point is the sum of the received power of the signals The delay spread capabilities of the various modulations are predominantly a function of the characteristics of that modulation. Given the delay spread capabilities of the various modulations, it is possible to predict system performance for the applicable modulation type and thereby design the system to meet coverage and propagation requirements. 17

DMR Simulcast Delay Spread - 3 Slave Master Slave Slave The whole simulcast-coverage area can be separated into two main types: capture areas: radio terminals receive a signal from only one transmitter (or where one signal only captures the receiver), non-capture (or overlap) areas: where the signal strength from adjacent transmitters is approximately equal In non-capture (or overlap) areas the delay made by the multipath from the two or more RBS toward the radio terminal must be simulated, controlled and adjusted to be non-destructive. 18

DMR Simulcast Delay Spread - 4 Hess described a method where multipath spread and the total signal power necessary for given BER criteria are plotted and used to determine coverage. BER (%) vs Multi-Path Spread (μs) The delay spread curves can be represented for various DAQ values, as the value of Tm is associated to the BER% and to the DAQ. TIA reported DAQ Scores DMR 2 slot TDMA (AMBE +2) 3.0 3.4 4.0 Simulcast Performance Cf/N vs Multi-Path Spread (μs) BER (%) 2.6 2.0 1.0 The loss of sensitivity becomes extreme as the delay spread increases. An important reference is the asymptote on the delay spread axis, which is the point at which it becomes impossible to meet the BER criterion independently of the signal strength. Another important reference for a modulation is the signal strength necessary for a given BER at Tm = 0 μs. Given these attributes, the delay performance of the candidate modulation is bounded. 19

DMR Simulcast Delay Spread - 5 Local reflectors Master Base Station Local reflectors Mobile d1 Slave Base Station d2 The delay spread can be expressed in a simplified way as: T ( d d ) 1 2 c 1000 20

DMR Simulcast Delay Spread - 6 Launch time adjustment: The usual procedure is to delay the transmission of RBS that have a smaller coverage area versus others which have a bigger coverage area In this example, the transmission of BS 3 must be delayed according to: ( 1 2 d d ) 1000 c = t 21

DMR Simulcast Delay Spread in practice - 1 If the differences in travel times of the various paths are negligible compared to the symbol time, the signal components arrive at the receiver at the same time, leading to an increase in the intensity of the signal received by the sum of the components The higher the bit rate of the system, the lower the symbol time => the differences in arrival times can become significant compared to the symbol time itself. This effect creates destructive Inter Symbol Interference (ISI) due to the sum of the current symbol with echoes of previous symbols, with an adverse effect on system performance The ISI begins to degrade the BER when the delay spread is greater than 1/7 to 1/10 of symbol time DMR modulation has a symbol duration of 208.3 µs, so attention has to be payed when the delay spread is higher than 20 to 30 µs 22

DMR Simulcast Delay Spread in practice - 2 To reduce the interference in the overlapping areas it is necessary to maintain the RMS delay spread at a value of less than 20 to 30 µsec To achieve this goal there are typically three means: by introducing a delay or an advance on the transmitted signal from RBS, this is achieved by changing Overlap Area Distance Compensation (OADC) parameters by changing the antenna bearing and kind (omnidirectional vs directional antenna) by changing the RF power level of the RBS (also with potential energy savings) The delay spread has higher impact in outdoor environments because multiple paths have more impact than in indoor environments 23

DMR Simulcast Delay Spread in practice - 3 Delay spread analysis should be performed within a distance range great enough (ideally infinity) to take into account also the RF field components from those RBS which, while not contributing to the communication (far more than the theoretical 75Km for DMR), can still contribute the delay spread Once completed, the static and dynamic analysis and solved the delay spread issue, it is necessary to verify that the new coverage allows the communication in all the user s area. This is because the changes due to the solution of the delay spread issue change the RF coverage of the system If needed, it is necessary to iterate the described procedure starting from the calculation of the link budget and correct the parameters (powers, delays, antenna bearings and types, etc.) so as to ensure RF coverage and solve the delay spread issue in the service area 24

DMR Simulcast Delay Spread - simulation examples 1 Performance for DAQ = 3.4 (BER = 2.0%) delay offset for equalized map The Signal Delay Spread (SDS) capability is the amount of delay between two independently faded equal amplitude signals that can be tolerated, when the standard input signal is applied through a faded channel simulator that will result in the standard BER at the receiver detector. The channel simulator provides a composite signal of two, equal amplitude, independently faded rays, the last of which is a delayed version of the first. 25

DMR Simulcast Delay Spread - simulation examples 2 26

Thank you! For more info: alessandro.guido@leonardocompany.com 27

I M P R E S S U M PMRExpo 2017 28. bis 30. November 2017 in Köln www.pmrexpo.de Veranstalter und Herausgeber EW Medien und Kongresse GmbH Kleyerstr. 88 60326 Frankfurt am Main www.ew-online.de November 2017 Copyright: Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt vor allem für Vervielfältigungen in irgendeiner Form (Fotokopie, Mikrokopie oder ein anderes Verfahren), Übersetzung und die Einspeicherung und Verarbeitung in elektronischen Systemen.