University of Toronto ECE 464 Winter 2021

ECE 464 January 26, 2021

L #6
Instructor: Ivo Maljević

5G link budget
Link budget is an important tool in a cellular network design. It is used to determine the cell radius or, in other words,
coverage area. By knowing the cell radius, we know how closely spaced cellular sites need to be. For instance, if
we have a 10km×10km area, and if we determine that sites need to be located every 1km (individual cell radius is
500m), we need approximately 100 sites to cover the en re area. If, on the other hand, sites need to be located ev-
ery 250m, we need approximately 40×40 = 1600 sites, which implies a much higher cost of the network deployment.

We start the 5G link budget exrcise by first defining the link budget as a measure of Maximum Allowable Pathloss
(MAPL, or simply MPL) between the transmi er and the receiver both in the downlink and uplink direc on. From the
MAPL, and by making use of one of the pathloss models introduced in the lectues, we can determine the cell radius.
A 5G link budget diagram is illustrated below.

Entries with a black font such are conducted Tx radio power, basesta on antenna height hBS and user terminal heights
hUT are provided as input parameters, although in prac ce they may also be part of the link budget evalua on (i.e.,
higher basesta on antenna height will result in higher MAPL). Power levels incicated with a blue font, like the EIRP
value and the UT receiver sensi vity, are calculated and so is the cell radius corresponding to the calculated MAPL
value. Losses (red font) and gains (green fonts) are part of the calcula on. The MAPL value, before various margins
are applied, is simply the difference between the EIRP level and the receiver sensi vity:
MAPL = EIRP + GUT − PUT sensi vity

and if we include the clu er (shadowing margin), interference margin and body loss, the MAPL value is reduced, re-
sul ng in shorter cell radius, as illustrated above.

Page 1 of 5
University of Toronto ECE 464 Winter 2021

Let us first look at the downlink 4G LTE link budget example, for a carrier opera ng at fc = 2150 MHz. The output
power at the radio port, called conducted power, is 46 dBm (40 W). The cable that connects the radio to the antenna
is a few tens of meters long, and reduces the signal power by 2 dB, whereas the transmit antenna gain is 18 dBi (see
the illustra on below).

An illustra on of a physical antenna mounted on top of a building with jumper cables connec ng it to the radio is
shown below.

The EIRP value, in dBm scale, is:

EIRP = 46 − 2 + 18 = 62 dBm.
In this example, the target data rate at the cell edge is set to 1 Mbps, which is quite low by today's standards (typically,

Page 2 of 5
University of Toronto ECE 464 Winter 2021

10 Mbps or more are targeted). Assuming a 10 MHz channel bandwidth (the occupied bandwidth is W = 9 MHz),
and α = 0.65 is the Shannon capacity scaling factor, the required SNR for the desired service is:
1
SNR = 2 α×W − 1 = 0.12579

which corresponds to SNR |dB = −9 dB.

To calculate the sensi vity of the user terminal (UT), we need the noise figure, which in this example is NF = 7 dB, so
that the sensi vity level is:

PUT sensi vity = −174 + 10 log10 (W ) + NF + SNR |dB

( )
= −174 + 10 log10 9 × 106 + 7 − 9
= −106.46 dBm.

The MAPL before the margins is:

MAPL = EIRP + GUT − PUT sensi vity

= 62 + 0 − (−106.46)
= 168.46 dB.

Win an interference margin of IM = 5 dB, the maximal allowable pathloss before clu er and penetra on losses is:

MAPL = 168.46 − 5 = 163.46 dB.

To determine the cell radius for non-line of sight condi ons (NLOS), that is when the UT is indoors and the basesta on
is outdoors, we need to add the following margins:
1. Shadowing margin - based on the shadowing standard devia on σ = 6 dB for urban environment (see TS
38.901) and a typical target of 85% cell edge reliability, the shadowing margin is calculated as:
√
M = 2σerfcinv (2POUTAGE ) = 6.22 dB

2. Penetra on loss - for a given carrier carrier frequency of fc = 2150 MHz, which corresponds to mid-bands
in terms of frequency ranges for LTE, wall penetra on is typically between 17 and 20 dB. We will assume a
penetra on loss of 18 dB.
3. Body loss - typical value is 3 dB when a user terminal is held in hand.
A er combining all the margins, the reduced MAPL is:

MAPL = 163.46 − 6.22 − 18 − 3 = 136.24 dB.

By inver ng the Urban Macro formula for NLOS condi on from the TS 38.901, we can calculate the cell radius for the
given carrier frequency and pathloss:

P LNLOS (d3D ) = 13.54 + 39.08 log10 (d3D ) + 20 log10 (fc ) − 0.6 (hUT − 1.5)
136.27 = 13.54 + 20 log10 (2.15) + 39.08 log10 (d3D )
39.08 log10 (d3D ) = 116.08123080168791
116.08
d3D = 10 39.08 = 934m
√
2
d2D = CellRadius = d23D − (hBS − hU T ) = 933.7m.

Note: The TS 38.901 specifies the pathloss for hBS = 25m only. For different basesta on antenna heights, refer to TR
36.873. Also, see the Python notebook for the pathloss implementa on that is based on two different specifica ons.

Page 3 of 5
University of Toronto ECE 464 Winter 2021

Using the same methodology outlined above, but taking into account the impact of the control plane overhead, we
can calculate the 5G link budget. The only difference compared to above is in the mapping from SNR to spectral
efficiency, which when the overhad is taken into account, is:
Rdata 1
SE =
BWN 1 − OHCP
Task, 5G link budget: For the 5G link budget calcula ons, we will assume the following:
• Carrier frequency: fc = 2150 MHz

• Channel bandwidth: BW = 40 MHz. For the noise bandwidth, assume 216 resource blocks, which translates
to BWN = 216 × 0.18 = 38.88 MHz.
• Conducted transmit power: PT = 200 wa s (53 dBm)
• Cable loss: Lcable = 0.5 dB

• Tx antenna gain (basesta on side): GT = 18 dBi

• Rx antenna gain (at the UT): GR = −1 dBi
• UT noise figure: NF = 7 dB
• Alpha-Shannon scaling factor: α = 0.6 (note: normally, mapping between the SE and the required SNR is done
using vendor specified tables, but for the purpose of evalua on, the TR 38.803 formula can be used)
• Control plane overhead: OHCP = 0.3, that is, 30%
• Pathloss model: Urban Macro, non-line of sight (NLOS), following 3GPP TS 38.901. This model assumes hBS =
25m and hUT = 1.5m.

• Interference margin: IM = 4 dB
• Body loss: Lbody = 3 dB
• Building penetra on loss: BPL = 20 dB
• Cell edge reliability: 85% (POUTAGE = 0.15)

• Target cell edge data rate: Rdata = 20 Mbps

Solu on: Each of the steps is outlined below.

1. The EIRP value at the transmi er, in dBm scale, is:

EIRP = dBm.

2. Receiver sensi vity at the UT is calculated through the following steps:

The receiver noise floor is;

PN = −174 + 10 log10 (BWN ) + NF = dBm

The spectral efficiency required to achieve a target data rate is:

Rdata 1
SE = = bps/Hz
BWN 1 − OHCP

Page 4 of 5
University of Toronto ECE 464 Winter 2021

and the required SNR is:

SE
SNR = 2 α − 1 =
SNR |dB = 10 log10 (SN R) = dB.

Finally, the receiver sensi vity is:

Psensi vity = PN + SNR |dB

=
= dBm.

The minimal UT signal level, taking into account the antenna gain of the receiver and the interference margin
is:
PUT,min = Psensi vity − GR + IM = dBm.

3. Maximal pathloss, before the margins are applied, is:

MAPL = EIRP − PUT,min = dB.

4. All the margins combined are:

Marginstotal = BPL + Lbody + M = dB,

where M = dB is previously calculated shadowing margin.

5. Maximal pathloss, a er the margins are applied, is:

MAPL = − Marginstotal = dB.

6. The cell radius is determined by inver ng the pathloss model from the TS 38.901, the same way as in 4G exam-
ple. Calculated distance (see the Python code) is:

CellRadius = m.

Page 5 of 5