QoS-aware energy-saving mechanism for hybrid optical-wireless broadband access networks

Abstract

Hybrid optical-wireless broadband access network (HOWBAN) takes full advantage of the high capacity and reliability of the passive optical network and the flexibility, ubiquity of the wireless network. Similar to other access networks, the issue of high energy consumption is a great challenge for HOWBAN. In HOWBAN, optical network units (ONUs) consume a great amount of energy. The sleep of ONUs can greatly improve the energy efficiency of HOWBAN. However, the quality of service (QoS) will be decreased while the packets are waiting in ONUs and optical line terminal. In this paper, we propose a QoS-aware energy-saving mechanism. A dynamic bandwidth allocation mechanism is designed to guarantee the QoS, where different priorities are considered. Meanwhile, we employ different sleep strategies by taking different priorities’ tolerant delays into account to prolong the sleep time of ONUs. Then, based on the evaluation of packet delay, the optimal sleep parameter is derived to maximum the energy efficiency. In addition, a load balancing and resource allocation mechanism is adopted in the wireless domain to reduce the delay and congestion caused by ONUs’ sleep. Results show that the proposed mechanism can effectively improve the energy efficiency and meet the QoS requirements of packets.

Appendix

The detailed process of the DGM (1, 1) is as follows:

Step 1 Accumulated generating operation

The sequence of the previous n slots’ average loads for ONU i is expressed as follows:

$$\begin{aligned} X^{\left( 0 \right) }=\left\{ {x^{\left( 0 \right) }\left( 1 \right) ,x^{\left( 0 \right) }\left( 2 \right) ,\ldots ,x^{\left( 0 \right) }\left( n \right) } \right\} \end{aligned}$$

(33)

Making one accumulated generating operation (1-AGO)

$$\begin{aligned} X^{\left( 1 \right) }=\left\{ {x^{\left( 1 \right) }\left( 1 \right) ,x^{\left( 1 \right) }\left( 2 \right) ,\ldots ,x^{\left( 1 \right) }\left( n \right) } \right\} \end{aligned}$$

(34)

where $x^{\left( 1 \right) }\left( k \right) =\sum \nolimits _{i=1}^k {x^{\left( 0 \right) }} \left( i \right) ,\quad k=1,2,\ldots ,n$.

1-AGO makes any nonnegative, fluctuant and nonfluctuant column translate to nondecreasing and increasing column, which weakens the randomness and strengthens the regularity.

Step 2 Establishment of DGM (1, 1)

The DGM (1, 1) mode is established as follows:

$$\begin{aligned} x^{\left( 1 \right) }\left( {k+1} \right) =\beta _1 x^{\left( 1 \right) }\left( k \right) +\beta _2 \end{aligned}$$

(35)

Step 3 Solutions of parameters

Let $\hat{{\beta }}=\left[ {\beta _1 ,\beta _2 } \right] ^{T}$ be the parameter column, and

$$\begin{aligned} Y=\left[ {\begin{array}{l} x^{\left( 1 \right) }\left( 2 \right) \\ x^{\left( 1 \right) }\left( 3 \right) \\ \vdots \\ x^{\left( 1 \right) }\left( n \right) \\ \end{array}} \right] ,\quad B=\left[ {\begin{array}{ll} {x^{\left( 1 \right) }\left( 1 \right) }&{}\quad {1} \\ {x^{\left( 1 \right) }\left( 2 \right) }&{}\quad {1} \\ {\vdots }&{}\quad {\vdots } \\ {x^{\left( 1 \right) }\left( {n-1} \right) }&{}\quad 1 \\ \end{array}} \right] \end{aligned}$$

(36)

Then, we can get

$$\begin{aligned} \hat{{\beta }}=\left( {B^{T}B} \right) ^{-1}\,B^{T}Y=\left[ {\begin{array}{l} \beta _1 \\ \beta _2 \\ \end{array}} \right] \end{aligned}$$

(37)

Substituting Y and B into Eq. (37), $\beta _1 ,\beta _2$ are given by

$$\begin{aligned} \beta _1= & {} \frac{\sum \nolimits _{k=1}^{n-1} {x^{\left( 1 \right) }\left( {k+1} \right) } \cdot x^{\left( 1 \right) }\left( k \right) -\frac{1}{n-1}\cdot \sum \nolimits _{k=1}^{n-1} {x^{\left( 1 \right) }\left( {k+1} \right) } \cdot \sum \nolimits _{k=1}^{n-1} {x^{\left( 1 \right) }\left( k \right) } }{\sum \nolimits _{k=1}^{n-1} {\left[ {x^{\left( 1 \right) }\left( k \right) } \right] ^{2}-\frac{1}{n-1}\left[ {\sum \nolimits _{k=1}^{n-1} {x^{\left( 1 \right) }\left( k \right) } } \right] } ^{2}}\nonumber \\ \end{aligned}$$

(38)

$$\begin{aligned} \beta _2= & {} \frac{1}{n-1}\cdot \left[ {\sum \limits _{k=1}^{n-1} {x^{\left( 1 \right) }\left( {k+1} \right) -\beta _1 \sum \limits _{k=1}^{n-1} {x^{\left( 1 \right) }\left( k \right) } } } \right] \end{aligned}$$

(39)

Step 4 Acquirement of prediction equation

Substituting $\beta _1 ,\beta _2$ into Eq. (35), then the recurrence function of DGM (1, 1) is obtained as follows

$$\begin{aligned} \hat{{x}}^{\left( 1 \right) }\left( {k+1} \right) =\beta _1^k x^{\left( 0 \right) }\left( 1 \right) +\frac{\left( {1-\beta _1^k } \right) \beta _2 }{1-\beta _1 },\quad k=1,2,\ldots ,n-1 \end{aligned}$$

(40)

The prediction equation is obtained by inverse 1-AGO as

$$\begin{aligned} \hat{{x}}^{\left( 0 \right) }\left( {k+1} \right) =\hat{{x}}^{\left( 1 \right) }\left( {k+1} \right) -\hat{{x}}^{\left( 1 \right) }\left( k \right) ,\quad k=1,2,\ldots ,n-1 \end{aligned}$$

(41)

where $\hat{{x}}^{\left( 0 \right) }\left( {k+1} \right)$ is the predicted value of ONU i’s average load in the next slot.