Multi-users network model and the corresponding networking scheme for indoor VLC systems

Tao Shang; Tao Jiang; Yin Tang Yang; Ping Wang; Yu Liu

doi:10.1364/OE.23.011600

1. Introduction

Recently, LEDs technology has been improved significantly and it has been recognized as the dominant lighting source in the future because of its advantages such as high luminance, long lifetime, low cost, small size, low power consumption, high modulation speed, etc. LEDs can be switched very quickly, and the idea of using high-speed flashing and dimming signal emitted by LEDs to transmit data has been put forward. In 2000, KEIO research group of Japan presented the basic thought that white LEDs can be used for wireless home link [1], and in 2004, the basic theory and channel model of visible light communication (VLC) were proposed [2].

Compared with other wireless communication technologies, like Wi-Fi and Bluetooth, VLC has inherent advantages of large capacity, unlicensed spectrum, immunity to radio frequency, high security, low cost, etc. It has been proposed as a supplement to Wi-Fi to provide high-speed access for phones, laptops and other indoor devices, especially in some occasions that radio signals are prohibited, such as in the flight. At present, VLC has attracted increasing attention [3–5], and it has made some progress in modulation bandwidth [6], multiplexing [7], and layout scheme of the light sources [8]. In indoor VLC systems, the different paths and reflections from the walls will cause multi-path effect to degenerate the communication performance. OFDM cannot only improve the capacity of system, but also eliminate the multi-path effect, and it has been widely researched in VLC systems [9,10].

To realize indoor VLC network, the reasonable full-duplex mode is significant. Using visible light for uplink and downlink [11], it is difficult to assure the communication quality because of the mutual interference among the links. The current way to avoid mutual interference is to use infrared for uplink and visible light for downlink. However, multiple terminals integrated with LEDs are necessary to afford the illumination and communication in usual interior space. In the situation of multiple terminals, some issues should be solved such as topological structure, network model, multi-users access method, and routing strategy. Heejin Lee et al. proposed that the VLC network can be established through the PLC (power line communication) and the Ethernet interface [12]. IEEE 802.15.7 standard expounds three modulation schemes and the corresponding theories in detail and also gives the fundamental statements about mobility support and the Media Access Controller (MAC) protocol [13].

In this paper, we adopt the hybrid full-duplex of infrared for uplink and visible light for downlink. In order to make sure that the user communicates in different directions, an illumination/communication terminal and the corresponding network model are proposed.

Normally, the uplink and downlink of the user are managed by one terminal. When multiple users connect to one terminal via infrared uplink, the access collision probability increases. Even if the access collision is solved, the network will be congested in the situation that multiple users transmit information though one terminal at the same time. Therefore, the access scheme is necessary to avoid the access collision and network congestion. In a multipoint transmission system, the user in the VLC coverage of a certain terminal could choose other terminals to access the network without unnecessary movement, which could avoid the access collision and network congestion to a certain extent. In this way, the uplink and downlink of the user can be probably managed by different terminals. Namely, the uplink of the user is managed by one terminal and the downlink is managed by one of the others. Thus the user access is accomplished by two terminals together and the information should be transmitted among terminals. However, the uplink-managed terminal does not know which terminal manages the downlink of the user, so the user should offer its “position” when accessing the system, and then the information forwarding can be performed. In this paper, in order to assure the multi-users access the VLC system effectively, we propose a multi-access scheme inspired by IEEE 802.15.7, which can solve the access collision and network congestion. We also present how to establish a communication link between users in detail and the routing strategy of indoor VLC system.

The rest of the paper is organized as follows: Section II describes the illumination/communication terminal and indoor VLC network model. Section III proposes the multi-access scheme, and analyzes the mobility support. Section IV presents the method of establishing a communication link between users and the routing strategy. Section V discusses how to ensure the communication performance when the user locates in the overlapping area of the footprint of the LEDs. Section VI gives the simulation and analysis. Section VII expounds some important conclusion.

2. Network model based on illumination/communication terminal

2.1 Illumination/communication terminal

Figure 1 is a conceptual representation of illumination/communication terminal. The terminal is composed of two level structures. On the lower surface of the cylindrical structure, LEDs are integrated in all direction; in the small hemispheric structure, a number of infrared receivers are integrated in a circle. For practical applications, in order to achieve uniform luminance distribution, the number of LEDs should be determined by the room size. On the other hand, the number of the infrared receivers is dependent on its FOV (field of view) and the size of the terminal. Theoretically, when the users access the network, the more the receivers are integrated, the smaller the collision probability is.

Fig. 1 Schematic diagram of illumination/communication terminal and three infrared receivers are spaced distributed in 120°.

Download Full Size | PDF

2.2 Indoor VLC network model

The network model is shown in Fig. 2, and the dash line denotes infrared uplink. Without loss of generality, we take four terminals as an example to show the indoor VLC network model. A certain terminal connects external router and all the terminals connect each other through infrared (or by wired backbone). All users connect the terminals to transmit information though infrared and receive information though visible light.

Fig. 2 Schematic diagram of indoor VLC network model.

Download Full Size | PDF

In Fig. 2, taking User 3 as an example, we describe how the user effectively chooses the terminal to access the network. In case 1, when User 3 is in the VLC coverage of Terminal A, it can choose Terminal A as its uplink-managed terminal. In case 2, when User 3 is in the VLC coverage of Terminal C, because Terminal C has already managed the uplinks of several users, in order to prevent the access collision and the network congestion, User 3 have to choose other terminals (The best one is idle Terminal A) to access the network. Thus, the uplink of User 3 is managed by Terminal A and the downlink is managed by Terminal C. In order to access the VLC network successfully, the complete routing information of User 3 should be built. However, Terminal A doesn’t know which terminal manages the downlink of User 3, so User 3 should forward its downlink information (i.e. “my downlink-managed terminal is Terminal C”) to Terminal A, and then the information is exchanged between Terminal A and C. The corresponding multi-access scheme is proposed hereinafter.

3. Multi-access scheme in VLC network

In our proposed VLC network model, the uplink and downlink of the user can be managed by different terminals, which avoid the access collision and the network congestion to a certain extent. But for multi-users, the access collision still exits. In RF wireless communication, the multi-access method of CSMA/CA is used to avoid access collision. Different from RF wireless communication, when a user transmits data via infrared, other users can’t sense the infrared channel so that they don’t aware whether the channel is idle, so these users need some information about when they can access the network.

In order to assure multi-users access the indoor VLC network effectively, we propose a multi-access scheme which can avoid the access collision and relieve the network congestion.

In indoor VLC network, all the channel information of uplinks (CIU) is broadcasted to the user via visible light, so that the user can choose which terminal to connect. The CIU includes two kinds of information: one indicates whether the channel is idle or how long the channel keeps busy, and the other indicates how many users connect a terminal. CIUs of all channels are broadcasted in every DIFS (Distributed Inter-Frame Space). When User S enters the VLC network, it will receive CIUs of all channels through visible light, and it can choose an appropriate uplink (assuming that the uplink is managed by Terminal A). When the uplink is idle, User S can access the network via this uplink. When the time that the uplink keeps busy is over, User S will receive ACK and start the backoff algorithm. The rest procedure is expounded as follows.

1) As shown in Fig. 3(a), User S sends RTS^#1 to Terminal A.
2) After receiving the RTS^#1, Terminal A puts the address of User S into its routing list. Then, Terminal A sends down the CTS^#1 attached its own MAC address via visible light, as shown in Fig. 3(b), Simultaneously, Terminal A informs the other terminals that the channel has been occupied, and also informs the occupied time and the number of users. Then, other terminals send this information to their own VLC coverage area.
3) If User S is in the VLC coverage of Terminal A, it receives CTS^#1 (including the MAC address of Terminal A). Then User S transmits an answer message, RTS^#2, to tell Terminal A which terminal manages the downlink of User S. Here, RTS^#2 includes the MAC address of Terminal A. Thus, Terminal A obtains the complete routing information of User S’s uplink and downlink. And then, Terminal A sends the CTS^#2 to permit User S to transmit data. As a result, User S completes the access progress. If Terminal A doesn’t receive RTS^#2 within time T1, which means that User S is not in the VLC coverage of Terminal A, Terminal A should broadcast RTS^#1 to its neighbor terminals, B and D.
4) After receiving the broadcasting RTS^#1, Terminal B and D send down the CTS^#1 attached their MAC address via visible light, as shown in Fig. 3(c).
5) When User S receives the CTS^#1 sent by Terminal D, it will transmit RTS^#2 to tell Terminal A that it is in the VLC coverage of Terminal D. Here, RTS^#2 includes the MAC address of Terminal D. Then Terminal A will obtain the complete routing information of User S. As a result, User S completes the access progress. In other word, User S has established the infrared uplink and visible light downlink with Terminal A and D, respectively. If Terminal A doesn’t receive RTS^#2 within time T2, which means that User S is not in the VLC coverage of B or D. The notice of continuously forwarding the RTS^#1 is sent by Terminal A, and Terminal B and D will broadcast RTS^#1 to Terminal C.
6) When User S is in the VLC coverage of Terminal C, Terminal C will receive the RTS^#1 forwarded by Terminal B or D, as shown in Fig. 3(d). When User S receives the CTS^#1 sent by Terminal C, it will transmit RTS^#2 to tell Terminal A that it is in the VLC coverage of Terminal C (seen in Fig. 3(e)). Here, RTS^#2 includes the MAC address of Terminal C. Terminal A broadcasts CTS^#2 to Terminal C, and Terminal C sends this message to User S, as shown in Fig. 3(f). As a result, User S completes the access progress and can transmit data to Terminal A (seen in Fig. 3(g)).

Fig. 3 Diagram of User S access, RTS^#1 indicates that User S sends request message, RTS^#2 indicates that User S sends request message with MAC address of a certain terminal, CTS^#1 indicates that User S sends the clear message with MAC address of a certain terminal, CTS^#2 indicates that User S sends the clear message.

Download Full Size | PDF

RTS^#1 is considered to be collided when User S doesn’t receive CTS^#2 within the time T3. And then, User S starts collision backoff algorithm and waits for the next access. The flowchart of User S accessing VLC network is shown in Fig. 4.

Fig. 4 Flowchart of User S accessing VLC network.

Download Full Size | PDF

For the proposed multi-access scheme, its key performance indicators, the collision probability and system throughput, are given. The transmission probability of the package τ is expressed as [14]:

τ = \frac{2 (1 - 2 p)}{(1 - 2 p) (W + 1) + p W (1 - {(2 p)}^{m})}

where, p is the collision probability; W is the collision window which depends on the number of the failed sending packets. In the first transmission, the minimum collision window is set as W. Every collision, W doubles, until it reaches its maximum value.

W_{\max} = 2^{m} W_{\min}

, where W_min and W_max are decided by the physical layer, and m is the maximum number of backoff.

Assumed that the number of the users is n, and each user has information to transmit. Because there are three infrared receivers in one terminal, the collision probability is written as:

p = 1 - {(1 - \frac{1}{3} τ)}^{n - 1}

Because the propagation distance is quite short, the propagation delay can be ignored.

For n users, the probability that at least one data package is transmitted is written as:

p_{t r} = 1 - {(1 - τ)}^{n}

The probability that at least one data package is transmitted successfully is:

p_{s} = \frac{n τ {(1 - τ)}^{n - 1} + C_{n}^{2} \frac{2}{3} τ^{2} {(1 - τ)}^{n - 2} + C_{n}^{3} \frac{2}{3} \frac{1}{3} τ^{3} {(1 - τ)}^{n - 3}}{p_{t r}}

The time of successful access can be written as:

\begin{array}{l} T_{s u c c} = R T S^{# 1} / R a t e_{I R} + S I F S + C T S^{# 1} / R a t e_{V L C} + N \times (R T S^{# 1} / R a t e_{I R} + S I F S) \\ + C T S^{# 1} / R a t e_{V L C}) + S I F S + R T S^{# 2} / R a t e_{I R} + S I F S + C T S^{# 2} / R a t e_{I R} + S I F S \\ + (h e a d e r_{M A C} + h e a d e r_{P H Y}) / R a t e_{I R} + P a y l o a d / R a t e_{I R} + S I F S \\ + N \times A C K / R a t e_{I R} + A C K / R a t e_{V L C} + D I F S \end{array}

where, N is the forwarding number; SIFS is the Short Interframe Space; DIFS is the DCF Interframe Space; header_MAC and header_PHY are the load of MAC header and PHY header respectively; Rate_VLC and Rate_IR are the date rates for the up and down link respectively; and ACK is the Acknowledgment message. The time of access failure can be represented as:

T_{f a i l} = R T S^{# 1} / R a t e_{I R} + D I F S

The normalized throughput can be represented as:

S = \frac{p_{s} p_{t r} (H e a d e r + P a y l o a d) / R a t e_{I R}}{p_{s} p_{t r} T_{s u c c} + p_{t r} (1 - p_{s}) T_{f a i l}}

T1, T2, and T3 must be confirmed when the user accesses the VLC network. T1 is defined as the time interval that the uplink-managed terminal does not receive RTS^#2 and RTS^#1 is forwarded once. T2 is defined as the one that the uplink-managed terminal does not receive RTS^#2 and RTS^#1 is forwarded twice. T3 is defined as the time interval that the user does not receive CTS^#1 when it sends the RTS^#1. According to the procedure that the user accesses the VLC network, T1, T2 and T3 can be written as:

T 1 = (C T S^{# 1}) / R a t e_{V L C} + S I F S + (R T S^{# 2}) / R a t e_{I R}

T 2 = (R T S^{# 1} / R a t e_{I R} + S I F S + C T S^{# 1} / R a t e_{V L C} + S I F S + R T S^{# 2} / R a t e_{I R})

T 3 = 2 \times (R T S^{# 1} / R a t e_{I R} + S I F S + C T S^{# 1} / R a t e_{V L C} + S I F S + R T S^{# 2} / R a t e_{I R})

When User S moves, it needs to consider whether to access the network again. Several situations are discussed below.

1) User S is still in the VLC coverage of the previous downlink-managed terminal and its infrared uplink remains unchanged.
2) User S is still in the VLC coverage of the previous downlink-managed terminal. Due to User S is outside of the FOV of the previous infrared receiving element (i.e. the direction of the previous infrared receiving element does not match with the infrared emission angle of User S), its infrared uplink needs to change.
3) User S is outside of the VLC coverage of the previous downlink-managed terminal. Nevertheless, its infrared uplink remains unchanged.
4) User S is outside of the VLC coverage of the previous downlink-managed terminal, and its infrared uplink needs to change.

For these four cases, User S should choose an appropriate terminal to access the network again when the interruption of its infrared uplink may occur. The process of accessing the network will repeat. If User S can send messages but cannot receive any response, User S is beyond the VLC coverage of the previous downlink-managed terminal. It will send change downlink message (CDLM) to its current uplink-managed terminal. Then, through broadcasting the CDLM, the uplink-managed terminal can find the downlink-managed terminal of User S.

4. Link establishment method between users and routing strategy

User S1 and S2 have already accessed the indoor VLC network. If User S1 wants to connect User S2, the link between User S1 and User S2 should be established. Assumed that the uplink and downlink of User S1 is both managed by Terminal A, User S1 transmits the request message to Terminal A via the infrared uplink, which includes ID of User S1 and S2, service ID and QoS. Then, Terminal A will conform whether its routing list contains the routing record of User S2. The rest procedure of establishing the link between User S1 and S2 is expounded as follows:

1) If the routing list of Terminal A contains the routing record of User S2, it means that the uplink of User S2 is managed by Terminal A, as shown in Fig. 5(a). When the routing list of Terminal A shows that the downlink of User S2 is not managed by Terminal A, Terminal A will send a message packet of Host-to-Host (HtoH) to Terminal C which manages the downlink of User S2. Then, Terminal C sends down the request message to User S2, and User S2 will transmit a feedback message to Terminal A. Thus, the link establishment between User S1 and S2 is accomplished. In the process above, when the routing list shows that the downlink of User S2 is also managed by Terminal A, Terminal A sends down the request message directly without sending HtoH to other terminals.
2) If the routing list of Terminal A doesn’t contain the routing record of User S2, it means that the uplink of User S2 is not managed by Terminal A, as shown in Fig. 5(b). Terminal A forwards a message packet of HtoH adhered its MAC address to the adjacent terminals. These terminals check their routing lists. If the routing record of User S2 cannot be found, these terminals will continue to forward the HtoH with their MAC addresses. If the routing record of User S2 is found, the HtoH will be sent to the downlink-managed terminal of User S2. Then, User S2 sends the response message to its uplink-managed terminal. Thus, the complete routing information is obtained, and the link establishment between User S1 and S2 is accomplished.

Fig. 5 Establishing links between users in the VLC network.

Download Full Size | PDF

The flowchart of establishing the link between User S1 and S2 is shown in Fig. 6.

Fig. 6 Flowchart of establishing the link between User S1 and S2 in the VLC network.

Download Full Size | PDF

If the interior space is large, a lot of terminals are necessary. The topology of multiple terminals is shown in Fig. 7, here A, B, C, D, E, F, G, H, and I represent terminals. We present the following routing strategy to ensure the network efficiency.

Fig. 7 Topological structure of illumination/communication terminals.

Download Full Size | PDF

1) The uplink-managed terminal only needs to forward the access message (RTS^#1) to the adjacent and diagonal terminals.
2) To establish the link between the users, the request message is forwarded to the adjacent terminals and can’t be forwarded back. When a certain terminal can’t forward the request message to other terminals, it is regarded that the message is transmitted to the edge of the network, and this message does not have to be forwarded.
3) When the terminal receives the same message, just deal with one.

5. Layout design of illumination/communication terminals

When the user locates in the overlapping area of the footprint of the LEDs, the mutual interference from different terminals occurs, which will degenerate the communication performance.

The visible light communication of each terminal is in certain footprint, and the footprint of two or more terminals may overlap. We define this overlapping area as the VLC overlapping area. The user in such areas needs to distinguish the messages from different terminals, and receives its own one. In order to make full use of the visible light bandwidth to meet the requirements of the multi-users communication, the signal is orthogonally encoded before being transmitted (for example, the OCDMA technology can be adopted in the indoor VLC networks), and then the interference signals can be eliminated by decoding the orthogonal codes. But there is a special situation that the user cannot distinguish messages when it receives several messages with the same code. According to our proposed networking scheme, this situation only occurs when the user accesses the network in the VLC overlapping area of terminals on the diagonal. In this situation, the user will receive the same CTS^#1 from the terminals on the diagonal at the same time. So, we design the layout of the terminals. It cannot only meet the requirement of indoor uniform illumination, but also eliminate the VLC overlapping area between the illumination/communication terminals on their diagonal, which ensures the user receive its own information accurately.

In order to design the terminal layout, the illumination and radiation characteristic of LED are analyzed.

The illumination distribution of LED is shown in Fig. 8. Obeying Lambertian distribution, the light intensity at the emission angle θ is given by:

I (θ) = \frac{m + 1}{2 π} I \cos^{m} (θ)

where, I(θ) is the light intensity of the LED, m is the radiation pattern of the light source, which is determined by the half power angle of the LED, namely:

Fig. 8 Diagram of the LED illumination distribution.

Download Full Size | PDF

m = - \ln 2 / (\cos Φ_{1 / 2})

For the light link of direct incidence, the horizontal illuminance E of the receiving point P(x, y, z) is written as:

E = \frac{I (θ) \cos (φ)}{d^{2}}

where, d is the distance between the LED center and the receiving point P(x, y, z), φ is receiving FOV of the point P(x, y, z).

For the 1st diffuse link, we can divide the light transmission into two processes: one is from the LED center to Point P₁ on the metope; and the other is from Point P₁ to the receiving point P(x, y, z). According to Eq. (11)-(13), the illuminance E₁ of Point P₁ is expressed as:

E_{1} = \frac{I \cos^{m} (α) \cos (β)}{l_{1}^{2}}

In general, the metope reflection is regarded as a Lambert one. So, Point P₁ can be considered as a new Lambert light source with the same radiation pattern of LED, and, ρD₁ and ρ represent the illuminance of Point P₁ and the reflection coefficient of the metope respectively. Thus, the illuminance E_i from P₁ to P(x, y, z) can be expressed as:

E_{i} = ρ \frac{I \cos^{m} (α) \cos (β) \cos^{m} (γ) \cos (ψ)}{l_{1}^{2} l_{2}^{2}} d A

The total illuminance of the 1st diffuse link, E_ref, is the sum of the illuminance that directly reach point P(x, y, z) after diffuse reflection of each point on four metopes, namely:

E_{r e f} = \sum \int_{w a l l} d E_{i}

According to Eq. (13) and (16), the total illuminance of the receiving point P(x, y, z) can be obtained.

The radiation characteristic of LED is shown in Fig. 9. On the receiver, the power of the visible light has the following form:

P_{R} = H (0) P_{S}

where, P_s is the transmission power of the LED, H(0) is the DC gain of the channel, which is expressed as:

H (0) = {\begin{matrix} \frac{(m + 1) A_{R}}{2 π d^{2}} \cos^{m} (ϕ) \cos (ω), 0 \leq ω \leq ω_{c} \\ 0, ω > ω_{c} \end{matrix}

where, A_R is the receiving area of PD detector, ω is the incident angle, φ is the emission angle, and ω_c is the FOV of the receiver.

Fig. 9 Diagram of the LED radiation characteristic.

Download Full Size | PDF

Thus, the receiving light power of the direct link can be obtained:

P_{R} = {\begin{matrix} P_{S} \frac{(m + 1) A_{R}}{2 π d^{2}} \cos^{n} (ϕ) \cos (ω), 0 \leq ω \leq F O V \\ 0, ω > F O V \end{matrix}

The ratio of the direct light, the 1st reflected light, and the 2nd and more times reflected light to all the receiving optical power are 95.16%, 3.57%, and 1.27%, respectively [2]. This indicates that, the performance of the VLC system is mainly affected by the directed light. Furthermore, utilizing OFDM technology, the influence of the diffuse link can be eliminated when we analyze the received power.

According to the above analysis, we put forward a method to determine the terminal layout and eliminate the VLC overlapping area of terminals on the diagonal, as shown in Fig. 10. Taking terminal B and D as an example, we put the detector at the midpoint of the diagonal of B and D. Given the sensitivity of the detector and the transmission power, we can calculate the distance between the terminal and the center T, which means that the position of the terminal on the diagonal is determined. The receiving power of T is the minimum power that the detector can detect. When the user is outside of this range, it will not be able to receive the message from the terminal. Thus, the VLC overlapping area of the terminal B and D on their diagonal will be eliminated. The specific calculation is as follows.

Fig. 10 Diagram of VLC coverage of terminal, the VLC overlapping area between the terminals on their diagonal is eliminated.

Download Full Size | PDF

Without loss of generality, we place the detector in horizontal direction. Hence, the emission angle of the LED is equal to the receiving angle of the detector, set as ω. h is the vertical height from the LED array to the receiving plane, L is the distance between the terminal and point T, and P_R is the receiver sensitivity. So we can get:

P_{S} \frac{(m + 1) A_{R}}{2 π L^{2}} \cos^{m + 1} (ω) = P_{R}

L = h / \cos ω

Substituting Eq. (21) into Eq. (20), so:

\cos ω = \sqrt[m + 3]{\frac{2 π R h^{2}}{(m + 1) A_{R}}}

and

L = \frac{h}{\sqrt[m + 3]{\frac{2 π R h^{2}}{P_{s} (m + 1) A_{R}}}}

According to Eq. (23), the position and the VLC coverage communication of single terminal can be determined, which is described as a circular region in Fig. 10. And there is no VLC overlapping area between the terminal B and D on the diagonal.

6. Simulation and analysis

Referring to literature [2,15,16], the simulation parameters are as follows: the size of the room is 5☓5☓3 m³; the receiving height of the user is 0.85 m from the floor; the vertical distance between LEDs and the roof is 0.15 m; the vertical distance between LEDs and the user’s receiving height is 2 m; the luminous intensity of a single LED is 0.73 cd; the luminous power is 20 mW; the array of LEDs is 60☓60; the half-power angle of LEDs is set as 60°; the receiver sensitivity is −17 dBm; the effective receiving area is 15 mm²; and the FOV of the receiver is 60 degree. From Eq. (23), the calculated coordinates of the terminals are A(1.0351, 3.9649), B(3.9649, 3.9649), C(3.9649, 1.0351) and D(1.0351, 1.0351), respectively.

We intercept the optical power plane of the diagonal midpoint, at which the value is the receiving sensitivity of the detector. Thus, there is no VLC overlapping area of Terminal B and D on their diagonal, as shown in Fig. 11.

Fig. 11 Diagram of VLC overlapping area of Terminal B and D on their diagonal.

Download Full Size | PDF

According to coordinates of the terminals, the indoor illumination is simulated as shown in Fig. 12. The values are within 540.5514 to 907.4149 lx, which meets the requirement of International Office Lighting.

Fig. 12 Indoor illumination distribution.

Download Full Size | PDF

According to IEEE Standard 802.15.7 MAC specifications [17] (Tables and Figures in this paragraph are all cited from this literature), the size of PHY header is set as 64 bits (Table 81 and Table 82); the size of MAC header is set as 240 bits (Fig. 44); when the basic bit rate of a single LED is 4 Mbps, the values of aMaxPHYFrameSize and aMinMPDUOverhead are 65535 bits and 72 bits respectively (Table 99 and Table 59); the value of aMaxMACPayloadSize is calculated as 65463 bits (Table 59), so the Payload size is set as 16384 bits which also conforms to the IrDA Link Access Protocol [19]; the base slot duration is 60 optical clocks (Table 59) and the frequency of the optical clock is 7.5 MHz (Table 74), thus the base slot duration is 8 μs; the SIFS is 120 optical clocks (Table 77), viz., 16 μs; because $D I F S = S I F S + (2 * b a s e_s l o t_d u r a t i o n)$ , the DIFS is 32 μs. Referring to 802.11 MAC specifications [14,18], the sizes of ACK, RTS^#1, RTS^#2, CTS^#1 and CTS^#2 are set.

The parameters of indoor VLC network are listed in Table 1.

Table 1. Time parameters in indoor VLC network

View Table

From Eq. (1) and (2), the calculated results of the collision probability are shown in Fig. 13. It can be seen that, when the number of the users increases, although the collision probability will increase, the collision probability is still small. For example, when the user number reaches 10, the collision probability is only about 0.24. The reason is that each terminal contains three infrared receivers. With the number of the infrared receivers increasing, the collision probability will become much smaller.

Fig. 13 Collision probability of different number of users.

Download Full Size | PDF

From Eq. (5)-(10), the date rates for the up and down links will affect the performance of the VLC networking. For a single LED, the data rate is about 4 Mbps. When with reuse of volume for higher aggregate speed, tens of 10 Mbps can be realized [15]. Referring to literature [15,16,20], we set the rates of the infrared uplink as 2, 5, 10 Mbps respectively and the rate of the visible light downlink as 100 Mbps. From Eq. (7), the calculated results of the normalized network throughput are shown in Fig. 14. It can be seen that the main influence factors are the rate of the infrared uplink and parameter N which denotes the forwarding number of the message (RTS^#1) when the user accesses the VLC network. Taking uplink rate 10 Mbps as an example, when N = 0, the normalized throughput, S, is around 0.882, which means that the uplink and downlink of the user are managed by one terminal; when N = 1, S is around 0.853, which means that the adjacent terminal manages the downlink; when N = 2, S is around 0.826, which means that the diagonal terminal manages the downlink. We also show that high rate results in low efficiency since the time ratio of payload will be decreased by increasing the rate. In either case, the calculated result of normalized throughput is enough to illustrate that our multi-access scheme is effective.

Fig. 14 Calculated results of network normalized throughput.

Download Full Size | PDF

Without loss of generality, we take uplink rate 10 Mbps as an example to illustrate our proposed multi-users networking scheme hereinafter.

Meanwhile, the average access time is simulated, and the results are shown in Fig. 15. Every time a number of users have an arbitrary distribution in the room, and all of them access the network many times. By taking the mean of these values, it can be clearly seen that the average access time becomes long when the number of the users increases. Even though the number of the users reaches 10, the average access time is still less than 2.5 ms, which meets the requirement of the user access latency.

Fig. 15 Numerical simulation results of average access time.

Download Full Size | PDF

For the numerical simulation of the user’s link establishment time, assumed that User S1and S2 have accessed the network, the uplinks of User S1 and S2 are managed by Terminal A and B respectively, and User S1 is in the VLC coverage of Terminal A. User S2 moves arbitrarily once a minute and re-establish the link with User S1.

Fig. 16 and Fig. 17 show the location of User S2 and the link establishment time, respectively. The maximum link establishment time is only about 225 μs. It can be seen that the link establishment time may be the same when User S2 locates in different positions. If the forwarding path or the number of forwarding messages is the same, the link establishment time will be the same (ignoring the propagation delay indoors). For examples, for the fifth and seventh link (i.e. k = 5 and k = 7), though the downlink-managed terminal changes after User S2 moving, the number of forwarding messages is the same and, therefore, the link establishment time keeps unchanged. However, for k = 5 and k = 6, link establishment time is different. The reason is that, Terminal C and D manage the downlink of User S2 respectively after the fifth and sixth movement, which leads to different forwarding paths and different number of forwarding messages.

Fig. 16 Locations of User S2.

Download Full Size | PDF

Fig. 17 Result of link establishment time.

Download Full Size | PDF

7. Conclusion

Recently, VLC systems have attracted wide attention. However, in usual interior space, multiple terminals integrated with LEDs are necessary to afford the illumination and communication. In the situation of multiple terminals, some issues should be solved such as the topological structure, network model, multi-users access method and routing strategy etc.

In this paper, we adopt the hybrid full-duplex of infrared for uplink and visible light for downlink. However, the hybrid full-duplex is different from the traditional one, which permits the user connect other terminals which don’t “manage” the downlink of this user. Based on this hybrid full-duplex, in order to effectively realize multi-users indoor VLC system, we design an illumination/communication terminal and the corresponding network model. Inspired by IEEE 802.15.7, a multi-access scheme is proposed, which can avoid the access collision and network congestion. Meanwhile, we present a link establishment method between users, and also expound the routing strategy of information forwarding.

On this basis, the expressions of the network performance indicators are deduced, such as the access collision probability, the network throughput, the average access time and the link establishment time. The corresponding simulations are made. The access collision probability is only about 0.24 when the user number reaches 10. By setting the rates of the infrared uplink as 2, 5, 10 Mbps respectively and the rate of the visible light downlink as 100 Mbps, the normalized throughput is analyzed. When the uplink rate is 10 Mbps, if the uplink and downlink of the user are managed by one terminal, the normalized throughput is around 0.882; if the adjacent terminal manages the downlink of the user, the normalized throughput is around 0.853; if the diagonal terminal manages the downlink of the user, the normalized throughput is around 0.826. The average access time is less than 2.5 ms when the number of the users reaches 10. The maximum link establishing time is about 225 μs. The results show that our proposed multi-access scheme and routing strategy are feasible and effective for multi-users indoor VLC networks.

Acknowledgments

Authors would like to thank Dr. Xu long Shen for thoughtful suggestions during various phases of this work. Authors also thank the anonymous reviewers and the editor for their valuable comments. This work is funded by NSF grant number 61172080.

References and links

1. Y. Tanaka, S. Haruyama, and M. Nakagawa, “Wireless optical transmissions with white colored LED for wireless home links,” in Proceedings of IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (IEEE, 2000), pp. 1325–1329. [CrossRef]

2. T. Komine and M. Nakagawa, “Fundamental analysis for visible-light communication system using LED lights,” IEEE Trans. Consum. Electron. 50(1), 100–107 (2004). [CrossRef]

3. D. O’Brien, “Indoor optical wireless communications: recent developments and future challenges,” Proc. SPIE 7464, 74640B (2009). [CrossRef]

4. S. Haruyama, “Progress of visible light communication,” in Optical Fiber Communication Conference, collocated National Fiber Optic Engineers Conference (2010), paper OThH2.

5. Y. Q. Zheng and M. L. Zhang, “Visible light communications-recent progresses and future outlooks,” in Proceedings of Symposium on Photonics and Optoelectronic (2010), pp. 1–6. [CrossRef]

6. J. Vucicet, C. Kottke, S. Nerreter, K. Habel, A. Büttner, K. D. Langer, and J. W. Walewski, “230 Mbit/s via a wireless visible-light link based on OOK modulation of phosphorescent white LEDs,” in Optical Fiber Communication Conference, collocated National Fiber Optic Engineers Conference (2010), Paper OThH3.

7. A. M. Khalid, G. Cossu, R. Corsini, P. Choudhury, and E. Ciaramella, “1-Gb/s transmission over a phosphorescent white LED by using rate-adaptive discrete multitone modulation,” IEEE Photonics J. 4(5), 1465–1473 (2012). [CrossRef]

8. L. B. Zeng, D. O’Brien, H. L. Minh, G. E. Faulkner, K. Lee, D. Jung, Y. Oh, and E. T. Won, “High data rate multiple input multiple output (MIMO) optical wireless communications using white LED lighting,” IEEE J. Sel. Areas Comm. 27(9), 1654–1662 (2009). [CrossRef]

9. N. Saha, R. K. Mondal, N. T. Le, and Y. M. Jang, “Mitigation of interference using OFDM in visible light communication,” in Proceedings of International Conference on ICT Convergence (2012), pp. 159–162. [CrossRef]

10. M. Afgani, H. Haas, H. Elgala, and D. Knipp, “Visible light communication using OFDM,” in Proceedings of 2nd International Conference on Testbeds and Research Infrastructures for the Development of Networks and Communities (2006), pp. 129–134.

11. Y. Wang, Y. Wang, N. Chi, J. Yu, and H. Shang, “Demonstration of 575-Mb/s downlink and 225-Mb/s uplink bi-directional SCM-WDM visible light communication using RGB LED and phosphor-based LED,” Opt. Express 21(1), 1203–1208 (2013). [CrossRef] [PubMed]

12. H. Lee, Y. Kim, and K. Sohn, “Optical wireless sensor networks based on VLC with PLC-ethernet interface,” in Proceedings of World Academy of Science, Engineering and Technology (2011), pp. 225–228.

13. S. Rajagopal, R. Roberts, and S. Lim, “IEEE 802.15.7 visible light communication: modulation schemes and dimming support,” IEEE Commun. Mag. 50(3), 72–82 (2012). [CrossRef]

14. G. Bianchi, “Performance analysis of the IEEE 802.11 distributed coordination function,” IEEE J. Sel. Areas Comm. 18(3), 535–547 (2000). [CrossRef]

15. Z. Wu, “Free space optical networking with visible light: a multi-hop multi-access solution,” Diss. Boston Univ. (2012).

16. Y. H. Son, H. S. Kim, Y. Y. Won, and S. K. Han, “Bidirectional visible light wireless transmission supported by reflective semiconductor optical amplifier based optical access network,” Microw. Opt. Technol. Lett. 53(5), 1032–1036 (2011). [CrossRef]

17. “IEEE Standard for Local and Metropolitan Area Networks-Part 15.7: Short-Range Wireless Optical Communication Using Visible Light,” September 2011.

18. “IEEE Standard for Information Technology-Telecommunications and Information Exchange between Systems Local and Metropolitan Area Networks-Specific Requirements-Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications-Amendment 4: Enhancements for Very High Throughput for Operation in Bands below 6 GHz,” December 2013.

19. “IrDA Link Access Protocol (IrLAP) v.1.1,” June 1996.

20. Z. Wu and T. Little, “Network solutions for the line-of-sight problem of new multi-user indoor free-space optical system,” IET Commun. 6(5), 525–531 (2012). [CrossRef]

Multi-users network model and the corresponding networking scheme for indoor VLC systems

Abstract

1. Introduction

2. Network model based on illumination/communication terminal

2.1 Illumination/communication terminal

2.2 Indoor VLC network model

3. Multi-access scheme in VLC network

4. Link establishment method between users and routing strategy

5. Layout design of illumination/communication terminals

6. Simulation and analysis

7. Conclusion

Acknowledgments

References and links

Cited By

Figures (17)

Tables (1)

Equations (23)

Optics Express

Payload size	16384 bits
MAC header	240 bits
PHY header	64 bits
ACK	112 bits + PHY header
RTS^#1	160 bits + PHY header
RTS^#2	160 bits + PHY header + MAC header
CTS^#1	112 bits + PHY header + MAC header
CTS^#2	112 bits + PHY header
LED basic bit rate	4 Mbps
base slot duration	8 μs
SIFS	16 μs
DIFS	32 μs