1

An Overview of 3GPP Release-15 Study on

Enhanced LTE Support for Connected Drones
Siva D. Muruganathan†, Xingqin Lin†, Helka-Liina Määttänen†, Zhenhua Zou†, Wuri A. Hapsari‡, Shinpei Yasukawa‡
†
Ericsson, ‡NTT DOCOMO
Contact: [email protected]

the Release-15 study on enhanced LTE support for aerial

Abstract—Cellular connectivity to low altitude unmanned aerial vehicles in March 2017 [11]. This study assessed the
vehicles (UAVs) has received significant interest recently which performance of LTE networks supporting aerial vehicles with
has led to a 3GPP study on enhanced LTE support for connected
up to Release-14 functionality. The study was completed in
UAVs in Release 15. The objective of the study is to investigate the
capability of long-term evolution (LTE) networks for providing December 2017 and the outcomes are documented in the 3GPP
connectivity to UAVs. In this article, we provide an overview of the technical report TR 36.777 [1] including comprehensive
3GPP study. We first introduce UAV connectivity requirements analysis, evaluation, and field measurement results. With the
and performance evaluation scenarios defined in the study. We completion of this study item, 3GPP has started a follow-up
then discuss radio channel models and the key identified work item [12] to advance LTE technologies to provide more
challenges of using LTE networks to provide connectivity to
efficient cellular connectivity to aerial vehicles. The
UAVs. Finally, we summarize potential solutions to address the
challenges including interference detection and mitigation enhancements for enhanced LTE support for connected drones
techniques, mobility enhancements, and UAV identification. is expected to be specified by 3GPP at the conclusion of the
work item which is expected to conclude in June 2018.
I. INTRODUCTION In this article, we provide an overview of the 3GPP Release-
Most of the activities in terms of research and development 15 study on enhanced LTE support for aerial vehicles and
in the area of mobile broadband have been focused on providing summarize the key findings. The overview provided by this
wireless broadband communication to outdoor users on the article is an accessible first reference for researchers interested
ground or indoor users in buildings. Recently, the area of in learning the 3GPP state-of-the-art findings on cellular
providing cellular connectivity to low altitude unmanned aerial connected drones.
vehicles (UAVs, a.k.a. drones) has gathered increasing interest
from the industry [1]-[4], academia [5]-[6], and public policy II. PERFORMANCE REQUIREMENTS
makers [7]-[8]. There is a plethora of use cases for supporting Although aerial user equipments (UEs) can be deployed for
aerial vehicles on cellular networks, some of which include a plethora of use cases, the two main data types with regards to
search-and-rescue, surveillance, wildlife conservation, package wireless connectivity of aerial UEs are command and control
delivery, and monitoring of critical infrastructure [9]. data and application data.
When aerial vehicles are flying well above eNodeB antennas, The ability to send command and control traffic to aerial UEs
they may have a high likelihood of line-of-sight (LOS) from eNodeBs can significantly improve the safety and
propagation conditions to multiple neighbouring eNodeBs. In operation of aerial UEs. For instance, it is critical that
information such as changes in the flight route are conveyed to
such a scenario, an uplink signal transmitted from an aerial
the aerial UEs in a timely manner with sufficient reliability. In
vehicle may become visible and cause interference to multiple
the 3GPP Release-15 study, the performance requirements on
neighbouring eNodeBs [10]. If this interference is not command and control to ensure proper operational control of
controlled or mitigated, it may adversely impact the uplink aerial UEs were defined. In addition to defining requirements
performance of existing users on the ground. Hence, to protect for command and control, the Release-15 study also defined
the existing users in the LTE network, the network may have to requirements for application data. A summary of the agreed
perform certain actions such as ensuring that such interference performance requirements from the Release-15 study is given
is mitigated or performing admission control of aerial vehicles in Table 1.
in the network. As a prerequisite to either of these actions, the
network may first need to identify the aerial vehicles. With the III. EVALUATION SCENARIOS AND ASSUMPTIONS
abovementioned LOS propagation conditions, downlink signals To evaluate the performance of LTE networks in the
transmitted from multiple eNodeBs may cause downlink presence of LTE connected aerial vehicles, the following three
interference to the aerial vehicle. Another open issue is whether scenarios were defined in the Release-15 study [1]:
the existing mobility mechanism of LTE networks is sufficient 1. Urban-macro with aerial vehicles (UMa-AV)
or whether it needs enhancements to support cellular 2. Urban-micro with aerial vehicles (UMi-AV)
connectivity of aerial vehicles.
To better answer these issues and to understand the potential
of LTE networks for providing cellular connectivity to aerial
vehicles, the 3rd generation partnership project (3GPP) started
 2

Table 1. Performance Requirements [1]

Command and Control Data Application Data
• telemetry • video
• waypoint update for streaming
autonomous aerial vehicle • image transfer
Data type
operation • transmission of
examples
• real-time piloting other sensor data
• flight authorization
• navigation database update
one-way radio interface similar to LTE
Latency latency of 50 ms from eNodeB terrestrial UEs
to aerial UE
Uplink/downlink 60-100 kbps for both uplink up to 50 Mbps for
data rate and downlink uplink
Command and up to 10-3 packet error loss rate N/A
control
reliability

3. Rural-macro with aerial vehicles (RMa-AV)

UMa-AV represents scenarios where the eNodeB antennas
are mounted above the rooftop levels of surrounding buildings
in urban environment. Urban scenarios with below rooftop
eNodeB antenna mountings are represented by UMi-AV.
Figure 1. An illustration of the evaluation scenarios in the Release-
Larger cells in rural environment with eNodeB antennas
15 study on LTE connected aerial vehicles.
mounted on top of towers are represented by RMa-AV. Figure
1 illustrates the evaluation scenarios used in the Release-15
study along with the inter-site distance (ISD), building height, A. LOS Probability
and eNodeB height for each scenario. To define LOS probability, an aerial UE height dependent
In UMa-AV, UMi-AV and RMa-AV, aerial vehicles are modelling approach was adopted in the study. For all three
modeled as outdoor UEs with heights well above ground level scenarios, the LOS probability models defined in [13] are
(AGL). In the Release-15 study, a maximum height of 300 m reused for aerial UE heights below a lower height threshold.
AGL was considered for aerial UEs. For performance The height threshold is 22.5 m for UMa-AV and UMi-AV,
evaluations, the height of the aerial UEs was assumed to be while the height threshold is 10 m for RMa-AV.
uniformly distributed between 1.5 m AGL and 300 m AGL. Since eNodeB antennas are well above rooftop in UMa-AV
Fixed aerial UE heights of 50, 100, 200, or 300 m AGL were and RMa-AV, a 100% LOS probability is assumed above an
also considered in the study for system level performance upper height threshold. The height threshold is 100 m and 40
evaluations. m for UMa-AV and RMa-AV, respectively. As eNodeB
In addition to modelling aerial vehicles, the three scenarios antennas are below rooftop in UMi-AV, the probability of non-
model terrestrial users on the ground and inside buildings. In line-of-sight (NLOS) is generally higher in UMi-AV when
all three scenarios, the user distribution of both indoor and compared to UMa-AV and RMa-AV. Hence, the upper height
outdoor terrestrial users is modelled according to the existing threshold is not applicable in UMi-AV.
3GPP models defined in [13]. In the aerial UE height range between the lower and upper
The total number of UEs (including both aerial UEs and height thresholds, the LOS probability models in [1] were
terrestrial UEs) per cell is assumed to be 15. To study the derived via ray tracing simulations for UMa-AV and RMa-AV.
impact of supporting aerial UEs with different densities in a For UMi-AV, ray tracing simulations were used to derive a
cell, aerial UE ratios of 0%, 0.67%, 7.1%, 25%, and 50% were LOS probability model that applies for aerial UE heights above
considered. Note that in [1], aerial UE ratio is defined as the the lower height threshold. The details of the agreed LOS
ratio between the number of aerial UEs per cell and the number probability models can be found in Table B-1 of [1]. Figure 2
of terrestrial UEs (including both indoor and outdoor) per cell. illustrates the LOS probabilities derived in the Release-15 study
Further details of evaluation assumptions for system level for the three different scenarios at different aerial UE heights.
simulations and mobility simulations can be found in Annexes
B. Pathloss and Shadow-Fading
A.1 and A.2 of [1], respectively.
For all three scenarios, the pathloss and shadow-fading
IV. CHANNEL MODELLING models defined in [13] are reused for aerial UE heights below a
lower height threshold. The lower height threshold is 22.5 m
In UMa-AV, UMi-AV, and RMa-AV, the channel modelling for UMa-AV and UMi-AV, while the lower height threshold is
of terrestrial users is according to the channel models defined 10 m for RMa-AV.
in [13]. To characterize the channels between aerial UEs and In the aerial UE height range above the lower height
eNodeBs, the Release-15 study defined models for LOS threshold, pathloss and shadow-fading models for both LOS
probability, pathloss, shadow-fading, and fast-fading. In this and NLOS conditions were agreed in the Release-15 study
section, we highlight some key aspects of these models. considering field measurements and ray tracing simulation
results contributed by multiple sources. The detailed
 3

discussions which led to these agreed pathloss models can be aerial UEs, when they are airborne, experience LOS
found in [14]. The agreed pathloss and shadow-fading models propagation conditions to more cells with higher probability
can be found in Tables B-2 and B-3 of [1], respectively. Figure than terrestrial UEs. This generally translates into higher
3 shows the LOS pathlosses for the three scenarios at different interference caused by an airborne aerial UE to these cells. The
aerial UE heights. uplink interference over thermal (IoT) ratios for terrestrial UEs
is given in Figure 4(a) which shows the effect of increased
C. Fast-Fading
uplink interference on terrestrial UEs as the aerial UE ratio
Three alternative fast-fading models were agreed during the increases.
Release-15 study. The three alternatives differ in the angular Due to the increased uplink interference, the uplink
spreads, delay spreads, and K-factor ranges as well as throughput performance of terrestrial UEs degrades when the
modelling methodology. The first alternative is based on a aerial UE ratio is increased in the network. The degraded uplink
clustered delay line model which is derived according to the throughput of terrestrial UEs in turn increases the uplink
procedures outlined in Annex B.1.1 of [1]. The second resource utilization level in the network. In other words,
alternative (outlined in Annex B.1.2 of [1]) is based on aerial degraded uplink throughputs imply that the UEs take longer
UE height dependent modelling of angular spreads, delay time to transmit their data which will consume more resources
spreads, and K-factor. The third alternative (outlined in Annex and will lead to increased uplink resource utilization. An
B.1.3 of [1]) is based on the fast-fading model of [13] with the increased uplink resource utilization level inherently means an
K-factor set to 15dB. increased level of uplink interference in the network, which in
turn was observed to degrade the uplink performance of both
V. PROBLEMS IDENTIFIED DURING THE STUDY aerial UEs and terrestrial UEs. The corresponding uplink
During the Release-15 study, evaluations were performed results demonstrating the degraded uplink performance are
under the scenarios and channel models described in Sections given in Annexes D.2.1 and D.2.2 of [1].
III and IV, and interference problems were identified in both B. Downlink Interference
uplink and downlink for scenarios involving aerial UEs. In this
In the downlink, compared to a typical terrestrial UE, the
section, we highlight the uplink and downlink interference aerial UEs observe interference from more cells due to the LOS
problems identified during the study. propagation conditions experienced by aerial UEs when they
A. Uplink Interference are airborne. Figure 4(b) compares the five-percentile
downlink geometry experienced by the aerial UEs when
In the uplink, the aerial UEs were found to cause interference
compared to that experienced by terrestrial UEs. Here geometry
to more cells than a typical terrestrial UE could. This is because
hUE=1.5 hUE=25
1 1

0.8 0.8
LOS probability

LOS probability

0.6 0.6

0.4 0.4
UMa-AV UMa-AV
0.2 UMi-AV 0.2 UMi-AV
RMa-AV RMa-AV
0 0
1 2 3 1 2 3
10 10 10 10 10 10
2D distance [m] 2D distance [m]
hUE=50 hUE=100
1 1

0.8 0.8
LOS probability

LOS probability

0.6 0.6

0.4 0.4
UMa-AV UMa-AV
0.2 UMi-AV 0.2 UMi-AV
RMa-AV RMa-AV
0 0
1 2 3 1 2 3
10 10 10 10 10 10
2D distance [m] 2D distance [m]
hUE=200 hUE=300
1 1

0.8 0.8
LOS probability

LOS probability

0.6 0.6

0.4 0.4
UMa-AV UMa-AV
0.2 UMi-AV 0.2 UMi-AV
RMa-AV RMa-AV
0 0
1 2 3 1 2 3
10 10 10 10 10 10
2D distance [m] 2D distance [m]

Figure 2. Illustration of LOS probabilities derived in the Release-15 study for different aerial UE heights denoted by ℎ𝑈𝐸 in meters.
 4

is defined as the ratio of the average received power from the applying interference mitigation. In the context of aerial UEs,
serving cell to the sum of the average interference power and interference detection is also linked with flying mode
noise power. The average interference power is computed as recognition, as when the aerial UE is above certain height, both
the sum of average received power from all non-serving cells. uplink and downlink interference increase. The threshold here
The degraded downlink geometry experienced by the aerial depends on network deployment and scenario.
UEs is a result of receiving downlink inter-cell interference During the Release-15 study, potential solutions for
from multiple cells. interference detection were broadly categorized into either UE-
Due to the increased downlink interference, the downlink based or network-based solutions. In this section, we briefly
throughput performance of aerial UEs degrades. The degraded summarize these solutions described in Section 7.1 of [1].
downlink throughput of aerial UEs when coupled with
A. UE-based Solutions
increased aerial UE ratios increases the downlink resource
utilization level in the network. An increased downlink In an LTE network, UE can perform neighouring cell
resource utilization level inherently means an increased level of measurements such as reference signal received power (RSRP),
downlink interference in the network, which in turn degrades reference signal received quality (RSRQ), reference signal–
the downlink performance of both aerial UEs and terrestrial signal to interference plus noise ratio (RS-SINR). Downlink
UEs. The corresponding downlink results demonstrating the interference can be detected based on these UE measurements
degraded downlink performance are given in Annexes D.1.1 reported to the eNodeB. One key aspect is to link the triggering
and D.1.2 of [1]. of measurement reports to the changing interference conditions.
To address the challenges due to the uplink and downlink An enhanced triggering condition could be a function of more
interference, 3GPP studied interference detection and than single cell RSRP. For example, the measurement can be
mitigation techniques, mobility enhancements, and UAV triggered when multiple cell RSRP/RSRQ values are
identification. We overview these solutions in the remaining above/within a threshold or when a sum of RSRP/RSRQ values
parts of this article, with a summary being presented in Table 1. is above a threshold.
Uplink interference can be detected either based on
VI. POTENTIAL SOLUTIONS FOR INTERFERNCE measurements at eNodeB or based on measurements reported
DETECTION by the UE. Furthermore, maximum output power and physical
resource blocks (PRBs) utilized may also be useful.
In this section, we first discuss how to detect the Since these solutions are based on UE measurements,
uplink/downlink interference caused/observed by an airborne existing measurement reporting mechanisms can be enhanced
aerial UE. Interference detection is a useful prerequisite for to improve interference detection. Potential enhancements
UMa-AV LOS
-50
-60
-70
-80
Pathloss (dB)

-90
-100 hUE= 1.5 m
-110 hUE= 50 m
-120
hUE= 150 m
-130
-140 hUE= 300 m
-150
1 2 3
10 10 10
2D distance [m]
UMi-AV LOS
-50
-60
-70
-80
Pathloss (dB)

-90
-100 hUE= 1.5 m
-110 hUE= 50 m
-120
hUE= 150 m
-130
-140 hUE= 300 m
-150
1 2 3
10 10 10
2D distance [m]

RMa-AV LOS
-50
-60
-70
-80
Pathloss (dB)

-90
-100 hUE= 1.5 m
-110 hUE= 50 m
-120
hUE= 150 m
-130
-140 hUE= 300 m
-150
1 2 3
10 10 10
2D distance [m]

Figure 3. Illustration of LOS pathloss models derived in the Release-15 study for different aerial UE heights denoted by ℎ𝑈𝐸 in meters. Note
that a carrier frequency of 2 GHz is assumed for UMa-AV and UMi-AV, while a carrier frequency of 700 MHz is assumed for RMa-AV.
 5

(a) (b)

Figure 4. Illustration of UL and DL interference: (a) UL IoT of terrestrial UEs versus different Aerial UE ratios. Note that these UL IoT results
are drawn from Table D.3.1-2 of [1]. (b) Comparison of five-percentile geometry between aerial UEs and terrestrial UEs. Note that the results
presented above are the average of the results in Annex C.2 of [1].
include new triggering events, enhancements of triggering power control mechanism in LTE (see Clause 5.1 of [15] for
conditions, and the inclusion of more results in the details of the existing power control mechanism in LTE). With
measurement report. this technique, the aerial UEs can be configured with a different
Lastly, UE related information such as mobility history fractional pathloss compensation factor compared to that
reports, speed estimation, timing advance adjustment values configured to the terrestrial UEs. Note that depending on the
and location information were also found to be useful. aerial UE height, different aerial UEs can also be configured
with different fractional pathloss compensation factors. Based
B. Network-based Solutions
on Release-15 evaluations, it was observed that applying height
In the network-based solutions, interference detection is dependent fractional path loss compensation factors to aerial
performed via the exchange of information between eNodeBs. UEs can significantly improve terrestrial UE performance while
For example, measurements reported by UE can be exchanged yielding notable performance gains for the aerial UEs. Details
between eNodeBs. and evaluation results corresponding to this technique can be
Another example is exchanging uplink reference signal found in Section 7.3.2.1 and Annex F.1.1 of [1].
configuration information of aerial UEs. By exchanging this A second power control technique evaluated during the
information, a neighbouring eNodeB can measure the uplink Release-15 study is the use of UE specific P0 parameter (note
interference caused by an aerial UE via measuring the power of that P0 is an open loop power control parameter specified in
the uplink reference signal. LTE). With this technique, the aerial UEs can be configured
Information on eNodeB’s downlink transmission power can with a different P0 parameter compared to that configured to the
also be exchanged between eNodeBs. With neighbor eNodeB’s terrestrial UEs. Based on the Release-15 evaluations, it was
transmission power, the uplink pathloss between an aerial UE observed that applying a lower P0 parameter to aerial UEs can
and the specific target neighbor eNodeB can be determined improve terrestrial UE performance at the cost of reduced aerial
assuming reciprocity and from the UE’s measurement reports. UE performance. Since the P0 parameter can be UE specifically
The uplink interference then can be estimated from the configured in LTE, this technique does not require specification
transmission power and the uplink pathloss. enhancements. However, it was found that the range of values
It should be noted, however, that the feasibility of supported in LTE for UE specific P0 may need to be extended.
exchanging all this information depends on factors such as the Details and evaluation results corresponding to this technique
type of backhaul and the feasibility of exchanging such are given in Section 7.3.2.2 and Annex F.1.2 of [1].
information over a large number of eNodeBs. During the Release-15 study, closed loop power control
based technique was also evaluated. In this technique, the target
VII. UPLINK INTERFERNCE MITIGATION received power of the aerial UEs was adjusted considering
In this section, we summarize the various uplink interference measurement reports received from both serving and neighbor
mitigation techniques that were evaluated during the Release- cells. Based on the Release-15 evaluations, it was observed that
15 study. closed loop power control can result in mean throughput
performance gains for both terrestrial UEs and aerial UEs.
A. Uplink Power Control Details and evaluation results corresponding to this technique
UE specific fractional pathloss compensation is one of the are given in Section 7.3.2.3 and Annex F.1.3 of [1]. During the
power control techniques evaluated during the Release-15 Release-15 study, the need for closed loop power control to
study. As the name implies, this technique requires the cope with potential fast signal change in the sky was discussed.
introduction of a UE specific fractional pathloss compensation Such fast signal changes are possible in the sky given the aerial
parameter which is an enhancement to the existing open loop UEs may be served by the sidelobes of eNodeB antennas. To
 6

cope with such fast signal changes, the step size of the transmit since Release-13. Details and evaluation results of this
power control command may need to be increased which in turn technique are given in Section 7.2.2 and Annex E.1 of [1].
may require specification enhancements to the existing power
B. Directional Antennas at UE
control mechanism in LTE.
During the Release-15 study, other power control techniques In the Release-15 study, the use of directional antennas at the
requiring modifications to the power control mechanism in LTE aerial UE is shown to reduce the downlink interference to the
to consider interference from neighbor cells were also aerial UEs from a broad range of angles. From the evaluation
discussed. However, solutions based on these techniques were results in the Release-15 study, it was observed that with non-
not evaluated during the study. ideal LOS tracking at the aerial UEs, the mean downlink
terrestrial UE throughput performance loss can be limited when
B. FD-MIMO compared to the case when the aerial UEs are equipped with
With full dimensional multi-input multi-output (FD-MIMO), omnidirectional antennas. The mean aerial UE downlink
multiple receive antennas at the eNodeB can be used to mitigate throughput performance was also shown to be significantly
interference in the uplink. Based on Release-15 evaluations, it improved with this technique. This technique does not require
was observed that, FD-MIMO can limit the mean uplink any specification enhancements since the use of directional
terrestrial UE throughput performance loss when compared to antennas at the aerial UE is an implementation issue. Details
the case when FD-MIMO is not deployed at the eNodeB. As and evaluation results corresponding to this technique are given
FD-MIMO is supported in LTE since Release-13, this in Section 7.2.3 and Annex E.2 of [1].
technique does not require any specification enhancements.
C. Receive Beamforming at UE
Details and evaluations corresponding to this technique are
given in Section 7.3.3 and Annex F.2 of [1]. When the aerial UEs are equipped with more than 2 receive
antennas, receive beamforming can be an effective interference
C. Directional Antennas at UE mitigation technique in the downlink. From Release-15
During the Release-15 study, the use of directional antennas evaluations, it was observed that receive beamforming with 8
at the aerial UE was shown to reduce the uplink signal power receive antennas at the aerial UEs can improve mean downlink
from the aerial UE in a broad range of angles. This in turn helps throughput performance of all UEs notably. Specification
reduce the uplink interference caused by the aerial UEs. The enhancements are not needed for this technique as the
following types of aerial UE LOS tracking were considered: application of receive beamforming can be done via
• the antenna direction of the aerial UE is aligned with implementation. Details and evaluation results corresponding
the direction of travel (DOT) of the UE to this technique are given in Section 7.2.4 and Annex E.3 of
• the LOS direction to the serving cell is either ideally [1].
tracked by steering its antenna boresight towards the D. Intra-site JT CoMP
serving cell or non-ideally tracked UE with errors due
to practical constraints In the intra-site JT CoMP (joint transmission coordinated
Release-15 evaluation results showed that with the antenna multiple points) scheme, data are jointly transmitted to the UEs
direction of the aerial UE aligned with the DOT, the mean from multiple cells that belong to the same site. From Release-
terrestrial UE throughput performance loss can be limited when 15 evaluations, it was observed that intra-site JT CoMP can
compared to the case when the aerial UEs are equipped with improve mean downlink performance of all UEs at low offered
omnidirectional antennas. Furthermore, this technique was also traffic load when compared to the case when intra-site JT
shown to significantly improve the mean aerial UE throughput CoMP is not employed. Specification enhancements are not
performance. As the use of directional antennas at the aerial needed for this technique as intra-site JT CoMP can be already
UE is an implementation issue, this technique does not require supported by LTE. Details and evaluation results
any specification enhancements. Details and evaluation results corresponding to this technique are given in Section 7.2.5 and
corresponding to this technique are given in Section 7.3.4 and Annex E.4 of [1].
Annex F.3 of [1]. E. Coverage Extension
In the Release-15 study, coverage extension techniques were
VIII. DOWNLINK INTERFERNCE MITIGATION studied to enhance synchronization and initial access
In this section, we summarize the various downlink performance of aerial UEs. From Release-15 evaluations, it
interference mitigation techniques that were evaluated during was observed that the proportion of UEs achieving
the Release-15 study. synchronization and initial access can be improved via
coverage extension techniques. Specification enhancements
A. FD-MIMO
are not needed for this technique as coverage extension is
With FD-MIMO, multiple transmit antenna ports at the already supported in LTE since Release-13. Details and
eNodeB can be used to mitigate downlink interference to the evaluation results corresponding to this technique are given in
aerial UEs. Based on Release-15 evaluations, it was observed Section 7.2.6 and Annex E.5 of [1].
that FD-MIMO can limit the impact on mean terrestrial UE
throughput performance while providing downlink aerial UE F. Other Schemes
throughputs that satisfy the aerial UE throughput requirements In the Release-15 study, other schemes such as coordinated
discussed in Section II. This technique does not require any transmission of control and data from multiple cells was briefly
specification enhancements as FD-MIMO is supported in LTE discussed. It was concluded that the details of the specification
 7

Table 2. Summary of Issues and Potential Solutions

Issue Solution Specification Impact
Interference detection using existing UE measurement
Already supported in LTE up to Release-14 and no specification
reports such as RSRP, RSRQ, RS-SINR. Power headroom
enhancements needed.
reports may also be used for uplink interference detection.

Interference detection using enhanced measurement reporting

mechanisms such as definition of new events, enhancements Requires specification enhancements to define new events, enhanced
to triggering conditions and inclusion of further measurement triggering conditions, etc.
results in measurement report.
Interference
Detection Interference detection using UE based information such as
mobility history reports, speed estimation, timing advance No specification enhancements needed.
adjustment values and location information.
Interference detection via exchange of information between
eNodeBs. Examples of information that can potentially be
Specification impact may depend on the type of backhaul. For instance,
exchanged include the following: (1) uplink scheduling
with non-ideal backhaul, the exchange of target eNodeB’s downlink
information or uplink reference signal configuration, (2)
transmission power will need specification enhancements.
target eNodeB’s downlink transmission power, (3) UE
measurement reports such as RSRP, RSRQ, RS-SINR
The introduction of UE specific fractional pathloss compensation
Uplink power control schemes such as applying UE specific parameter requires specification enhancement. Similarly, for closed
loop power control, the introduction of increased step size of the
fractional pathloss compensation, applying UE specific P0 transmit power control command requires specification enhancement.
parameter, and applying closed loop power control with The application of UE specific P0 parameter does not require
Uplink increased step size of the transmit power control command.
Interference specification enhancement, although the range of values supported in
Mitigation LTE for UE specific P0 may need to be extended.
Already supported in LTE up to Release-14 and no specification
FD-MIMO
enhancements needed.
Directional Antennas at UE An implementation issue and no specification enhancements needed.
Already supported in LTE up to Release-14 and no specification
FD-MIMO
enhancements needed.
Directional Antennas at UE An implementation issue and no specification enhancements needed.
Receive Beamforming at UE An implementation issue and no specification enhancements needed.
Downlink Already supported in LTE up to Release-14 and no specification
Interference Intra-site JT CoMP
enhancements needed.
Mitigation
Already supported in LTE up to Release-14 and no specification
Coverage extension
enhancements needed.
Details of the specification impact depend on the details of the
Other Schemes coordinated data and control transmission scheme which needs further
study.
Enhancements to handover procedure such as conditional
handover and/or handover related parameters considering
Specification enhancements needed.
Mobility such as location information, airborne status, flight path
Performance plans, etc.
Improvement Enhancements to existing measurement reporting
mechanisms such as definition of new events, enhancements Specification enhancements needed.
of triggering conditions, etc.
Identifying a flying aerial UE based on such as flight mode
indication, altitude or location information, or implicitly via Specification enhancements needed.
enhanced measurement reporting.
Aerial UE
Identifying a flying aerial UE based on mobility history
Identification No specification enhancements needed.
reports/patterns.
Identifying an aerial UE based on subscription information in
combination with radio capability indication from the aerial Specification enhancements needed.
UE

impact depend on the details of the coordinated data and control

transmission scheme which needed further study. Evaluation IX. MOBILITY PERFORMANCE AND POTENTIAL
results corresponding to this technique are given in Section ENHANCEMENTS
7.2.7 and Annex E.6 of [1]. During the Release-15 study, mobility simulations were
performed (see Annex J of [1]) and measurements from field
trials were collected (see Annex H of [1]). From these results,
the mobility performance of an aerial UE is shown to be worse
 8

when compared to that of a terrestrial UE especially when the XI. CONCLUSIONS

number of aerial UEs in large. Due to the increased downlink In Release-15, 3GPP has dedicated a significant effort during
interference, the downlink signal to interference plus noise ratio its study on LTE connected drones and concluded that it is
(SINR) for the aerial UEs is much worse than the downlink feasible to use existing LTE networks to provide connectivity
SINR for the terrestrial UEs. Hence, the aerial UEs may to low altitude drones despite some challenges, as overviewed
experience more handover failures, more radio link failures, in this article. Providing efficient and effective connectivity to
longer handover interruption time, etc. The mobility simulation the aerial UEs while minimizing the impact on terrestrial
results showed a better mobility performance for aerial UEs in devices requires a rethinking of many of the assumptions,
the RMa-AV scenario than in the UMa-AV scenario. It should models, and techniques used to date for cellular system. This
be noted however that interference mitigation techniques listed article has particularly focused on the 3GPP state-of-the-art
in Sections VII and VIII were not considered in the mobility findings on LTE connected drones, although most of the lessons
simulations of the Release-15 study and the use of such herein would likely apply to any cellular systems (such as 5G)
techniques is expected to improve the aerial UEs’ mobility providing connectivity to the sky.
performance.
In the Release-15 study, the following techniques to improve REFERENCES
mobility performance of aerial UEs were identified:
[1] 3GPP TR 36.777, “Enhanced LTE support for aerial vehicles,” Online:
• enhancements to handover procedure such as ftp://www.3gpp.org/specs/archive/36_series/36.777.
conditional handover and handover related parameters [2] GSMA, “Mobile spectrum for unmanned aerial vehicles; GSMA public
considering such as location information, airborne policy position,” white paper, October 2017. Online:
status, flight path plans, etc. https://www.gsma.com/spectrum/wp-content/uploads/2017/10/Mobile-
spectrum-for-Unmanned-Aerial-Vehicles.pdf. Accessed on December
• enhancements to existing measurement reporting 14, 2017.
mechanisms such as new events, enhancements of [3] X. Lin, V. Yajnanarayana, S. D. Muruganathan, S. Gao, H. Asplund, H.-
triggering conditions, etc. L. Maattanen, M. Bergström, S. Euler, Y.-P. E. Wang, “The sky is not the
More detailed discussion can be found in Section 7.4 of [1]. limit: LTE for unmanned aerial vehicles,” IEEE Communications
Magazine, vol. 56, no. 4, pp. 204-210, April 2018. Available at
https://arxiv.org/abs/1707.07534.
X. AERIAL UE IDENTIFICATION [4] X. Lin, R. Wiren, S. Euler, A. Sadam, H.-L. Maattanen, S. D.
Muruganathan, S. Gao, Y.-P. E. Wang, J. Kauppi, Z. Zou, and V.
Depending on country-specific regulations, an aerial UE may Yajnanarayana, “Mobile Networks Connected Drones: Field Trials,
need to be identified by the network in order to allow the use of Simulations, and Design Insights,” submitted to IEEE Communications
LTE networks for aerial UE connectivity. Another aspect is that Magazine, February 2018. Available at http://arxiv.org/abs/1801.10508.
there may be drone specific service or charging by the operator. [5] M. Gharibi, R. Boutaba and S. L. Waslander, “Internet of drones,” IEEE
Access, vol. 4, pp. 1148-1162, March 2016.
In the Release-15 study, aerial UE identification was discussed; [6] S. Chandrasekharan, K. Gomez, and A. Al-Hourani, et al., “Designing and
see details in Section 7.5 of [1]. implementing future aerial communication networks,” IEEE
Aerial UE identification solution discussed during the study Communications Magazine, vol. 54, no. 5, pp. 26-34, May 2016.
item is a combination of user based identification via [7] FAA, “UAS traffic management research transition team plan,”
technical report, January 2017. Online:
subscription information and device functionality based https://www.faa.gov/uas/research/utm/media/FAA_NASA_UAS_Traffic
identification via LTE radio capability signaling. The mobility _Management_Research_Plan.pdf. Accessed on January 30, 2018.
management entity (MME) can signal the subscription [8] U.S. DOT and FAA, “UAS integration pilot program,” October 2017.
information to the eNodeB which can include information on Online:https://www.faa.gov/uas/programs_partnerships/uas_integration_
pilot_program/. Accessed on January 30, 2018.
whether the user is authorized to operate for aerial usage. In [9] Goldman Sachs, “Drones: Reporting for work,” 2017. Online:
addition, an aerial UE as LTE device can indicate its support of http://www.goldmansachs.com/our-thinking/technology-driving-
aerial related functions that will be introduced in the Release- innovation/drones/. Accessed on January 30, 2018.
15 work item [12] via radio capability signaling to the eNodeB. [10] V. Yajnanarayana, Y.-P. E. Wang, S. Gao, S. Muruganathan, X. Lin,
“Interference mitigation methods for unmanned aerial vehicles served by
The combination of the subscription information and the radio cellular networks”, IEEE 5G Word Forum, to appear. Available at
capability indication from the UE can be used by the eNodeB https://arxiv.org/abs/1802.00223
to identify an aerial UE, and then perform the necessary control [11] RP-170779, “Study on enhanced LTE support for aerial vehicles,” NTT
and the relevant functions. DOCOMO, Ericsson, March 2017. Online:
http://www.3gpp.org/ftp/tsg_ran/tsg_ran/TSGR_75/Docs/RP-
The LTE capability indication plays a role in the flight mode 170779.zip. Accessed on December 14, 2017.
recognition. An aerial UE capable of Release-15 enhancements [12] 3GPP RP-172826 “New WID on Enhanced LTE Support for Aerial
may be able to explicitly indicate flight mode if that is specified. Vehicles,” Ericsson, December 2017. Online:
Alternatively, the measurement triggering enhancements used http://www.3gpp.org/ftp/TSG_RAN/TSG_RAN/TSGR_78/Docs/RP-
172826.zip. Accessed on January 30, 2018.
for interference detection may be used to implicitly aid flight [13] 3GPP TR 38.901, “Study on channel model for frequencies from 0.5 to
mode detection. Flight mode detection of a UE that does not 100 GHz,” Online: ftp://www.3gpp.org/specs/archive/38_series/38.901
have Release-15 aerial UE capability needs to rely on existing [14] R1-1715084, “Summary of Email discussion [89-10] on remaining details
standardized metrics from UE such as mobility history, speed, of channel modelling”, Ericsson, August 2017. Online:
http://www.3gpp.org/ftp/tsg_ran/WG1_RL1/TSGR1_90/Docs/R1-
existing RSRP/RSRQ measurement events. In some networks, 1715084.zip. Accessed on January 4, 2018
this is also equivalent to rogue drone detection where a non- [15] 3GPP TS 36.213, “Evolved Universal Terrestrial Radio Access (E-
aerial UE is not allowed to use the network for connectivity UTRA); Physical layer procedures (Release 15),” V15.0.0. Online:
while airborne. ftp://www.3gpp.org/specs/archive/36_series/36.213/