Content uploaded by Yong Niu
Author content
All content in this area was uploaded by Yong Niu on Oct 08, 2017
Content may be subject to copyright.
QoS-aware Scheduling for Small Cell Millimeter
Wave Mesh Backhaul
Yun Zhu∗, Yong Niu†, Jiade Li∗, Dapeng Oliver Wu∗, Yong Li†and Depeng Jin†
∗Department of Electrical and Computer Engineering, University of Florida, Gainesville, FL 32611-6130, USA
†Tsinghua National Laboratory for Information Science and Technology (TNLIST),
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China
Abstract—With the explosive growth of mobile data demand,
small cells densely deployed underlying the homogeneous macro-
cells are emerging as a promising candidate for the fifth
generation (5G) mobile network. The backhaul communication
for small cells poses a significant challenge, and with huge
bandwidth available in the mmWave band, the wireless backhaul
at mmWave frequencies can be a promising backhaul solution
for small cells. In this paper, we propose the Maximum QoS-
aware Independent Set (MQIS) based scheduling algorithm for
the mmWave backhaul network of small cells to maximize the
number of flows with their QoS requirements satisfied. In the
algorithm, concurrent transmissions and the QoS aware priority
are exploited to achieve more successfully scheduled flows and
higher network throughput. Simulations in the 73 GHz band
are conducted to demonstrate the superior performance of our
algorithm in terms of the number of successfully scheduled flows
and the system throughput compared with other existing schemes.
I. INTRODUCTION
Mobile data demand is growing explosively. Some industry
and academic experts predict a 1000-fold demand increase
by 2020 [1]. In order to offer the 1000x increase in data
rates and throughput, small cells densely deployed underlying
the conventional homogeneous macrocells are emerging as
a promising candidate for the fifth generation (5G) mobile
broadband [2]. This new network deployment is usually re-
ferred to as heterogeneous cellular networks (HCNs). How-
ever, with the increase of the number of small cells deployed,
the backhaul for small cells becomes a significant challenge
[2], [3]. Although fiber based backhaul offers large bandwidth,
it is costly, inflexible, and time-consuming to connect the
densely deployed small cells. In contrast, wireless backhaul
is more cost-effective, flexible, and easier to deploy [3].
With huge bandwidth available, wireless backhaul in mmWave
bands, such as the 60 GHz band and E-band (71–76 GHz and
81–86 GHz), provides several-Gbps data rates and can be a
promising backhaul solution for small cells.
On the other hand, unlike existing communication systems
using lower carrier frequencies (e.g., from 900 MHz to 5
GHz), mmWave communications suffer from high propagation
loss. To combat severe channel attenuation, directional anten-
nas are utilized at both the transmitter and receiver for high
antenna gain. With the beamforming technique, the transmitter
and the receiver are able to direct their beams towards each
other for the directional communication [4]. The directional
communication reduces the interference between links, and
concurrent transmissions (spatial reuse) can be exploited to
greatly improve network capacity. In a scenario where small
cells are densely deployed, effective and efficient backhaul
scheduling schemes need to be designed with the characteris-
tics of mmWave communications taken into account.
0DFURFHOO
6PDOO&HOOV
*DWHZD\
(EDQG%DFNDKXO
Fig. 1. The mesh backhaul network in the small cells densely deployed
scenario.
In Fig. 1, we present a typical scenario of densely deployed
small cells underlying the macrocell cellular network. In the
small cells, mobile users are associated with the base stations
(BSs), and the BSs are connected via backhaul links with the
mesh topology. There are one or more BSs connected to the
backbone network via the macrocell site, which are called
gateways. In this targeted small cells system, the backhaul
network is in the E-band, which provides high data rates. For
the scheduling problem of the backhaul network for small
cells densely deployed, there are two aspects of challenges.
In the first aspect, concurrent transmissions need to be fully
exploited to maximize the spatial reuse gain. In the second
aspect, the scheduling scheme should provide the quality of
service (QoS) guarantee for flows in the backhaul network. To
ensure fairness, the scheduling scheme should maximize the
scheduled flows in the network with the QoS requirement of
each flow satisfied.
In this paper, we develop a QoS aware scheduling scheme
for the small cell backhaul network in the mmWave band. The
contributions of this paper are summarized as follows.
•We formulate the problem of optimal scheduling to max-
imize the number of flows with their QoS requirements
satisfied in the mmWave backhaul network as a nonlinear
IEEE ICC 2016 - Wireless Communications Symposium
978-1-4799-6664-6/16/$31.00 ©2016 IEEE
integer programming. Concurrent transmissions (spatial
reuse) are explicitly considered in this problem.
•We propose a heuristic scheduling algorithm to solve the
formulated problem with low complexity. The interfer-
ence between flows is modeled by the contention graph.
Based on the contention graph, we propose the Maxi-
mum QoS-aware Independent Set (MQIS) based back-
haul scheduling algorithm to achieve more successfully
scheduled flows and higher total network throughput.
•We evaluate our algorithm for the backhaul network in the
73 GHz band, and the realistic antenna model is adopted
in the simulation. The simulation results demonstrate the
superior performance of our algorithm in terms of the
number of successfully scheduled flows and the system
throughput compared with other existing schemes.
The structure of this paper is as follows. Section II describes
the related work. Section III gives an overview of the system
model. Mathematical analysis and problem formulation are
presented in Section IV. Section V presents the MQIS based
scheduling algorithm for the mmWave backhaul network. Ex-
tensive simulations are conducted and evaluated in Section VI.
Section VII concludes this paper.
II. RELATED WORK
Time division multiple access (TDMA) has been a promi-
nent solution for mmWave backhaul [5], [6]. Taori et al. [5]
proposed a time-division multiplexing (TDM) based schedul-
ing scheme to support point-to-multipoint, non-line-of-sight,
mmWave backhaul. Islam et al. [3] performed the joint cost
optimal aggregator node placement, power allocation, channel
scheduling and routing to optimize the wireless backhaul
network in mmWave bands. In [7], the scheduling for the radio
access and backhaul networks were jointedly designed. To the
best of our knowledge, none of the previous works are devoted
to address the balance between the QoS requirement and the
contention between links in the mmWave backhaul network.
On the other hand, similar problems have also been investi-
gated in WPANs [8], [9], [10]. One influential work is the
Exclusive Region (ER) based scheduling which is introduced
and derived in [8]. It ensures that concurrent transmissions
always outperform the serial TDMA by co-scheduling flows
in the exclusive region. Qiao et al. [10] proposed a concurrent
transmission scheduling with the QoS requirements of flows
considered. In Ref. [10] , the set of concurrent flows are chosen
in a greedy manner to maximize the overall system throughput,
through which the number of flows successfully scheduled is
maximized. However, existing scheduling approaches have not
fully utilized the global information of the contentions residing
in the network. And a more QoS-favorable strategy is desired.
In this paper, for the first time we introduce the concept of
QoS-aware independent set to the scheduling problem under
mmWave bands, and the proposed protocol achieves better
performance in terms of system throughput and the number
of scheduled flows in the mmWave backhaul network.
III. SYSTEM MODEL AND ASSUMPTION
A. Link Level QoS
In the backhaul network, a end-to-end flow with QoS
requirement may go through multiple hops with a proper
routing protocol in place. Once the routing path is fixed, all
the single-hop flows along the path will share the same QoS
constraint as the end-to-end traffic. Accordingly, for every
single hop link in the backhaul network, its QoS is defined as
the sum of QoS requirements of all the end-to-end traffic going
through it. To this end, the scheduling can only focus on the
single hop links with QoS requirements, and the designing of
routing protocol will be left for our future work. In following
paper, the word “flow” only means the single-hop link.
B. TDMA Structure
We consider the scenario where small cells are densely
deployed, and assume there is a backhaul network controller
(BNC) residing on one of the gateways. Each BS in the
network is equipped with an electronically steerable direc-
tional antenna, and can direct its beam towards other BSs for
directional transmission. In our investigated system, time is
partitioned into superframes, and each superframe consists of
Mtime slots called channel time allocation (CTA). We further
assume the transmission requests and signaling information
for mmWave backhauling are collected by the 4G BS by its
reliable transmission [6]. Thus the BNC is able to obtain the
transmission requests and the location information of other
BSs. In our scheme, with directional transmission, multiple
links can be scheduled concurrently in the same time slot,
which is also referred to as the spatial-time division multiple
access (STDMA) [10].
C. Physical Model
Since non-line-of-sight (NLOS) transmissions suffer from
higher attenuation than line-of-sight (LOS) transmissions [11],
we assume the directional LOS transmission between BSs can
be achieved with the locations of BSs adjusted appropriately
(e.g., on the roof). We assume there are Nflows requesting
transmission slots in the superframe, and each flow represents
one backhaul link. We denote the distance between the trans-
mitter siof flow iand the receiver rjof flow jby dij .We
also denote the antenna gain of siin the direction of from si
to rjby Gt(i, j), and the antenna gain of riin the direction
of from sjto riby Gr(j, i). Then considering the path loss
and signal dispersion over distance, the received power at the
receiver rifrom sican be calculated as
Pr(i, i)=k0Gt(i, i)Gr(i, i)d−n
ii Pt,(1)
where k0is a constant coefficient and proportional to (λ
4π)2(λ
denotes the wavelength), ndenotes the path loss exponent, and
Ptdenotes the transmission power [10]. Due to the half-duplex
assumption, adjacent links cannot be scheduled for concurrent
transmissions. If flow iand flow jare not adjacent, we denote
it by i∝j. Then under concurrent transmissions, the received
interference at rifrom sjcan be calculated as
Pr(j, i)=ρk0Gt(j, i)Gr(j, i)d−n
ji Pt.(2)
where ρis the multi-user interference (MUI) factor related to
the cross correlation of signals from different links.
According to the Shannon’s channel capacity, the achievable
data rate of flow ican be estimated as
Ri=ηWlog2(1 + Pr(i, i)
N0W+
j∝i
Pr(j, i)),(3)
where Wis the bandwidth, and N0is the onesided power
spectra density of white Gaussian noise [10]. η∈(0,1)
describes the efficiency of the transceiver design.
IV. PROBLEM FORMULATION AND ANALYSIS
In this section, we formulate the optimal scheduling problem
into a nonlinear integer programming problem.
We assume there is a minimum throughput requirement for
each flow i, and denote it by qi. We denote a schedule as
S, and assume it has Kstages. In each stage, multiple links
are scheduled for concurrent transmissions. For each flow i,
we define a binary variable ak
ito indicate whether flow iis
scheduled in the kth stage. If so, ak
i=1; otherwise, ak
i=0.
We denote the number of time slots of the kth stage by δk.
Since there are different links in different stages, we denote
the transmission rate of flow iin the kth stage by Rk
i. Then
we can obtain Rk
ias
Rk
i=ηWlog2(1+ ak
ik0Gt(i, i)Gr(i, i)d−n
ii Pt
N0W+ρ
j
ak
jk0Gt(j, i)Gr(j, i)d−n
ji Pt
).
(4)
Then we can obtain the throughput of flow ibased on Sas
Ti=
K
k=1
δk·Rk
i·tslot
t0+M·tslot
,(5)
where t0is the time duration of collecting transmission re-
quests and signaling information, and tslot is the time duration
of each time slot in the CTA period (CTAP). Then we define
a binary variable Qito indicate whether the QoS requirement
of flow iis satisfied in S. If so, Qi=1; otherwise, Qi=0.
Given the throughput requirements of flows, with the limited
number of time slots in the CTAP, the optimal schedule
should accommodate as many flows as possible. Therefore, the
optimal scheduling problem P1 can be formulated as follows.
(P1)max
N
i=1
Qi(6)
s.t.
Qi=1,if Ti≥qi,
0,otherwise; ∀i(7)
K
k=1
δk≤M;(8)
ak
i+ak
j≤1,if flow iand jare adjacent; ∀i, j (9)
This is a nonlinear integer programming problem, and is
NP-hard. Constraint (7) indicates if the throughput of flow iin
the schedule is larger than or equal to its throughput require-
ment, Qi=1; otherwise, Qi=0. Constraint (8) indicates
there are at most Mtime slots in the CTAP. Constraint (9)
indicates due to the half-duplex operation of BSs, adjacent
links cannot be scheduled for concurrent transmissions since
there is at most one connection for each node.
Since it is difficult to solve the problem of P1 in polyno-
mial time, we propose an efficient and practical scheduling
algorithm instead in the next section.
V. S CHEDULING ALGORITHM DESIGN
In this section, we propose the Maximum QoS-aware In-
dependent Set (MQIS) based scheduling algorithm for prob-
lem P1. As the name suggests, the Maximum QoS-aware
Independent Set is a combination of flows that has minimal
internal interference and is beneficial for QoS achievement.
In our algorithm, flows in this set are scheduled concurrently.
To present the scheduling algorithm, we first introduce the
contention graph under directional antennas, which captures
the global knowledge of interference; Then we define the QoS-
aware priority for each flow and present how to find a MQIS
with the contention graph and the priority value; Finally, we
give the overall MQIS based backhaul scheduling algorithm.
A. Contention Graph
The MQIS based scheduling summarizes the global inter-
ference information in the contention graph, in which a node
represents a real flow and an edge between a pair of nodes
marks the contention. We judge the existence of contention
between every pair of flows based on two principles: 1) the
half duplex nature where a single BS can not receive and
transmit packets at the same time. In other words, if two flows
share the same source or destination, there will be a contention
edge between them; 2) the impact that one flow has on another.
For every flow pair, we define the relative-interference (RI) as
follows:
RIj,i =Pr(j, i)
Pr(i, i)(10)
where Pr(j, i)and Pr(i, i)is defined by (2) and (1) re-
spectively. We insert an edge between flow iand flow jif
max(RIi,j ,RI
j,i)>σ, where σis a threshold.
B. QoS-aware Priority
We assign a priority value to each flow out of QoS consid-
erations. Flows that can achieve requested throughput more
quickly are preferred in our scheduling because they can soon
stop transmission and leave time slots for others to use. To
give more weight to those flows, we define the priority as the
inverse of the number of slots that a flow needs in CTAP to
achieve its QoS requirement. Based on previous definitions,
the priority of flow ican be expressed as follows:
priority(i)== Ri·tslot
qi∗(t0+M·tslot)(11)
Note that the data rate Rihere is computed without any
interference in (3), so the numerator represents how many bits
flow ican transmit in a single slot. The denominator in (11)
is the total amount of bits that need to be transmitted in one
super frame.
C. Find MQIS
In the MQIS based backhaul scheduling, the set of flows
scheduled at any slot should be a MQIS. Obviously when some
flows achieve QoS requirement and are removed from current
scheduling, this condition may no longer be satisfied. When
this happens, we will select flows from contention graph G
to add to the current scheduling set to generate a new MQIS.
To begin with, the “unqualified” flows which we will not se-
lect from should be removed from G. A flow is “unqualified”
if it satisfies one of the following conditions: 1) It has already
achieved QoS requirement so there is no need to consider it; 2)
It has been scheduled and thus ongoing now; 3) It is a neighbor
of one of the ongoing flows. The third condition comes from
the fact that neighbors in Gshould never appear together in
the MQIS.
Then, we iteratively select the best node (or flow) from the
remaining graph and add it to the scheduling set. First, we
hope the chosen node has few contented neighbors and thus
a more number of flows may be co-transmitted at the same
time slot. Recall that in the contention graph, for any node, the
larger degree means more contented neighbors, so we simply
choose the node that has minimal degree. However if there are
multiple nodes satisfying this criterion, we evaluate the other
aspect of those flows: the ability to achieve QoS requirement
quickly. We refer to the priority computed from (11) and
select the node with maximum priority value to add to the
scheduling set. After that, we remove the chosen node as well
as its neighbors in Gand begin to select the next one as long
as the remaining contention graph is not empty.
The detailed algorithm is summarized in Algorithm 1.
We use an array schedule to denote the scheduling set,
where schedule(i)=1means the flow iis scheduled and
schedule(i)=0means it is unscheduled yet. Moreover,
if flow ihas already achieved QoS requirement, we denote
schedule(i)=−1. In this algorithm, we use the existing
scheduling array as the input and generate a new one, which
is a MQIS. In line 5, we find the minimal degree set MD.
If there are more than one elements in MD, we select the
Algorithm 1 findMQIS
Input:
contention graph G
existing schedule array schedule
priority array priority
Output:
new scheduling array , schedule;
1: remove unqualified flows from G
2: while G=∅do
3: P= the remaining set of flows in G
4: calculate degree for flows in P
5: MD ={p|minp∈Pdegree(p)}
6: if |MD|>1then
7: i=max
i∈MD(priority(i))
8: else
9: choose i∈MD
10: end if
11: schedule(i)=1
12: find neighbor set Nof i
13: D={i}∪N
14: remove flows in Dfrom G
15: end while
16: return schedule
flow with maximum priority value in line 7. In line 12-14
we remove the newly scheduled flows and its neighbors in G
before starting the next round.
D. MQIS based Backhaul Scheduling
Now we summarize the overall scheduling algorithm for
the backhaul network. After the BNC receives QoS requests
from BSs, it will construct the contention graph Gand make
scheduling decisions. According to (4) (5), the slots can
be divided into a number of stages during which the same
scheduling is kept. In MQIS backhaul scheduling, the end of
one stage is the slot in which some scheduling flows have
achieved QoS requirement. We call those flows “finished”. In
other words, we should check at every slot if there are some
newly finished flows, and if so, a new MQIS should be found
using Algorithm 1.
For Nflows and Mslots in CTAP, we use a N∗Mbinary
matrix Bto denote the final scheduling S, where B(i, j)=1
means the flow iat slot jis scheduled. The detailed process is
shown in Algorithm 2. The initialization steps are among line
1-5. In line 8-11, we denote the newly finished flows in array
schedule as −1. Then, we will call Algorithm 1 to generate
the new scheduling array, as indicated by line 12.
VI. PERFORMANCE EVALUATION
In this section, we evaluate the performance of the proposed
scheduling algorithm in the 73GHz band, and compare it with
existing schemes.
A. Simulation Setup
We consider a backhaul network with 10 base stations which
has at most 90 flows. Since the scheduling performance is
Algorithm 2 MQIS Backhaul Scheduling Algorithm
1: BNC receives transmission request ri(i=1,2, ...N )
requiring minimum throughput Ri
min
2: construct contention graph G
3: initialize N∗Mscheduling matrix B=0
4: initialize schedule =(0,0, ...0) with length N
5: compute priority for flows
6: for slot k(1 ≤k≤M)do
7: if k=1or some flows newly finished then
8: denote the set of the newly finished flows as FIN
9: for f∈FIN do
10: schedule(f)=−1
11: end for
12:
schedule =findMQIS(G,schedule,priority)
13: B(:,k)=schedule
14: else
15: B(:,k)=B(:,k −1)
16: continue
17: end if
18: end for
19: return B
dependent on the location of stations, we randomly generate
a position for each BS within a 1000 square meters area.
Meanwhile, for every flow, we randomly choose its source
and destination. And the requested throughput for this flow
is uniformly distributed between 1 Gbps and 3 Gpbs. For the
path loss, we use the channel model in Ref [12]. Besides, we
adopt the widely used realistic directional antenna model in
Ref [13]. All the BSs in the system use the same transmission
power level. Some other parameters are shown in Table I.
TABLE I
SIMULATION PARAMETERS
system bandwidth W 1.2 GHz
transmission power Pt30 dBm
background noise N0-134 dBm/MHz
slot time tslot 18 μs
beacon period t0850 μs
Number of slots in transmission period M 2000
We implemented the serial TDMA, and the state-of-the-
art protocol STDMA [10] for comparison. To evaluate our
proposed protocol, the following metrics are considered:
•Number of successful flows: the number of flows that
achieve the required QoS. Note that if a flow has been
scheduled during the scheduling phase but can not satisfy
the QoS during the transmission phase, it will not be
counted as a successful flow.
•System throughput: the achieved total throughput of
the backhual network. In other words, this metric is the
average of sum of the throughputs of all flows.
Number of demanding flows
10 20 30 40 50 60 70 80 90
Number of successful flows
4
6
8
10
12
14
16
18
TDMA
STDMA
MQIS
(a) Number of successful flows
Number of demanding flows
10 20 30 40 50 60 70 80 90
System throughput(Gbps)
10
15
20
25
30
35
TDMA
STDMA
MQIS
(b) System throughput
Fig. 2. Performance under different number of flows
B. Comparison with Other Schemes
1) Under different number of flows: In this case, we choose
the number of slots in CTAP as 2000, and set σ=10
−4.We
vary the number of flows in the backhaul network from 10
to 90. With the increasing number of demanding flows, we
evaluate the two metrics and plot the results in Fig 2.
From the results, we can observe the trend of the perfor-
mance of the MQIS based scheduling algorithm under the
increasing number of demanding flows. The more demanding
flows there are, the more chances for the spatial reuse, and thus
both the number of successful flows and the system throughput
keep increasing. Due to the system capacity constraint, they
gradually become flat and reaches the capacity.
Compared with TDMA and STDMA, the MQIS based
scheduling has obvious advantages. TDMA has no spatial
reuse at all so it can only schedule limited flows. When
only a few flows are to be scheduled, the difference between
STDMA and MQIS is trivial because both schemes can almost
accommodate all the demanding flows. But as the number
of demanding flows increases, the MQIS based scheduling
can achieve better performance in two aspects. First of all,
when the number of demanding flows is around 10 to 20,
the performance of STDMA has already entered the flattened
phase where more number of flows will not bring obvious
better performance; However, the proposed scheme keeps
increasing dramatically until the number of demanding flows
Number of slots in CTAP
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Number of successful flows
4
6
8
10
12
14
16
18
TDMA
STDMA
MQIS
(a) Number of successful flows
Number of slots in CTAP
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
System throughput(Gbps)
5
10
15
20
25
30
35
TDMA
STDMA
MQIS
(b) System throughput
Fig. 3. Performance under different number of slots
reaches 80. Moreover, when the traffic demand is large, the
MQIS based scheduling can achieve around 60% more number
of successful flows and about 40% more system throughput
than STDMA.
The better performance of MQIS based scheduling algo-
rithm comes from two facts. First, it uses global contention
knowledge to make scheduling. For STDMA, a new flow
will be added to scheduling set as long as it can increase
the total throughput. This method may get stuck to bad
local optimal, where highly contented flows are co-scheduled.
While in MQIS scheduling, we always schedule the flows
that are relatively independent with each other, and thus more
close to the global optimal. Secondly, the QoS of a flow is
considered as an priority in MQIS based scheduling algorithm,
and contributes to the overall performance.
2) Under different number of slots: In this case, we aim to
compare the performance of different protocols under different
number of slots in CTAP. The number of demanding flows
is kept to be 90. We change the number of slots in CTAP
from 500 to 5000, and evaluate the two metrics as before. The
results are shown in Fig 3. As we can observe, the number of
successful flows and system throughput only slightly increase
as the number of slots in CTAP changes. With enough time
slots, the MQIS based algorithm can achieve 17 successful
flows while STDMA can only schedule around 11 flows.
Besides, the system throughput of the MQIS based algorithm
is 10 Gbps higher than that of STDMA.
VII. CONCLUSION
In this paper, we consider the problem of optimal scheduling
to maximize the number of flows with their QoS requirements
satisfied in the mmWave backhaul network. We have proposed
the MQIS based backhaul scheduling algorithm, where the
spatial reuse is fully exploited based on the contention graph.
The QoS aware priority is also exploited in the algorithm to
provide better QoS guarantees for flows. Extensive simulations
show our algorithm is able to achieve more successfully
scheduled flows as well as higher network throughput than
other schemes.
ACKNOWLEDGMENT
This work was supported in part by NSF CNS-1116970.
REFERENCES
[1] “Qualcomm data challenge,” accessed March 2015,
https://www.qualcomm.com/documents/rising-meet-1000x-mobile-
data-challenge.
[2] R. Baldemair, T. Irnich, K. Balachandran, E. Dahlman, G. Mildh, Y.
Seln, S. Parkvall, M. Meyer, and A. Osseiran, “Ultra-dense networks in
millimeter-wave frequencies,” IEEE Communications Magazine, vol. 53,
no. 1, pp. 202–208, January 2015.
[3] M. N. Islam, A. Sampath, et. al., “Wireless backhaul node placement for
small cell networks,” in 2014 48th Annual Conference on Information
Sciences and Systems (CISS) (Princeton, USA), Mar. 19–21, 2014, pp. 1–
6.
[4] J. Wang, Z. Lan, C. Pyo, T. Baykas, C. Sum, M. Rahman, J. Gao,
R. Funada, F. Kojima, and S. Kato, “Beam codebook based beamform-
ing protocol for multi-Gbps millimeter-wave WPAN systems,” IEEE J.
Selected Areas in Communications, vol. 27, no. 8, pp. 1390–1399, Oct.
2009.
[5] Taori, Rakesh, and Arun Sridharan. ”Point-to-multipoint in-band
mmwave backhaul for 5G networks.” Communications Magazine, IEEE
53.1 (2015): 195-201.
[6] J. Qiao, X. S. Shen, J. W. Mark, Q. Shen, Y. He, L. Lei, “Enabling
device-to-device communications in millimeter-wave 5G cellular net-
works,” IEEE Communications Magazine, vol. 53, no. 1, pp. 209–215,
January 2015.
[7] Niu, Yong, et al. ”Exploiting Device-to-Device Communications in Joint
Scheduling of Access and Backhaul for mmWave Small Cells.” (2015).
[8] L. X. Cai, L. Cai, X. Shen, and J. W. Mark, “REX: a Randomized
EXclusive Region based Scheduling Scheme for mmWave WPANs with
Directional Antenna,” IEEE Trans. Wireless Commun., vol. 9, no. 1,
pp. 113–121, Jan. 2010.
[9] J. Qiao, L. X. Cai, and X. Shen, et. al., “Enabling multi-hop concurrent
transmissions in 60 GHz wireless personal area networks,” IEEE Trans-
actions on Wireless Communications, vol. 10, no. 11, pp. 3824–3833,
Nov. 2011.
[10] Qiao, Jian and Cai, Lin X and Shen, Xuemin and Mark, Jon W,
“STDMA-based scheduling algorithm for concurrent transmissions in
directional millimeter wave networks,” IEEE International Conference
on Communications (ICC), pp. 5221–5225, 10-15 June 2012.
[11] S. Y. Geng, J. Kivinen, X. W. Zhao, and P. Vainikainen, “Millimeter-
wave propagation channel characterization for short-range wireless com-
munications,” IEEE Trans. Veh. Technol., vol. 58, no. 1, pp. 3–13, Jan.
2009.
[12] Akdeniz, Mustafa Riza and Liu, Yuanpeng and Samimi, Mathew K and
Sun, Shu and Rangan, Sundeep and Rappaport, Theodore S and Erkip,
Elza, “Millimeter wave channel modeling and cellular capacity evalu-
ation,” IEEE Journal on Selected Areas in Communications, vol. 32,
pp. 1164–1179, June 2014.
[13] Toyoda, I., T. Seki, and K. Iiguse, ”Reference antenna model with side
lobe for TG3c evaluation.” IEEE document, pp. 802–15, 2006.
