CN103327518B - Node dormancy scheduling method based on link quality - Google Patents

Node dormancy scheduling method based on link quality Download PDF

Info

Publication number
CN103327518B
CN103327518B CN201310253355.7A CN201310253355A CN103327518B CN 103327518 B CN103327518 B CN 103327518B CN 201310253355 A CN201310253355 A CN 201310253355A CN 103327518 B CN103327518 B CN 103327518B
Authority
CN
China
Prior art keywords
node
delay
link
time slot
reviving
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CN201310253355.7A
Other languages
Chinese (zh)
Other versions
CN103327518A (en
Inventor
陈良银
王金磊
张靖宇
孙利民
朱红松
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Institute of Information Engineering of CAS
Original Assignee
Institute of Information Engineering of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Institute of Information Engineering of CAS filed Critical Institute of Information Engineering of CAS
Priority to CN201310253355.7A priority Critical patent/CN103327518B/en
Publication of CN103327518A publication Critical patent/CN103327518A/en
Application granted granted Critical
Publication of CN103327518B publication Critical patent/CN103327518B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Mobile Radio Communication Systems (AREA)

Abstract

The invention relates to a node dormancy scheduling method based on link quality. The scheduling method comprises the step 1, detecting the link quality between adjacent nodes on a link, the step 2, combining the detected link quality to calculate a delay expextation value from a source node to a destination node, and carrying out the step 3 if the delay expectation value is larger than a delay threshold required by practical application or finishing procedures, and the step 3, from the beginning of increasing one revival time slot, sequentially calculating the minimum delay expectation value from the source node to the destination mode after the revival time slots are added until the calculated minimum delay expectation value is smaller than or equal to the delay threshold or times for increasing the revival time slots are equal to the hop count between the source node to the destination node. The node dormancy scheduling method takes the link quality into consideration, enables the delay between the ends of the node to be reduced by inserting time slots, and can meet limited delay of a system.

Description

A kind of node dormancy dispatching method based on link-quality
Technical field
The present invention relates to wireless sensor network technology field, relate to the dormancy dispatching side of wireless sensor network interior joint Method, be a kind of in given link by insert minimum time slot of reviving make link meet node dormancy that call duration time requires is adjusted Degree method.
Background technology
In order to overcome problem energy life time of wireless sensor network, energy research is always the weight of sensor network research Point.Have built up a lot of model in recent years for the environment around node is collected energy, thus it is powered.But Current research great majority concentrate on hardware designs and power management aspect, save energy by concept transfer duty-cycle The research of aspect is the most few.The proximal segment time, ESC agreement (Yu Gu, Ting Zhu, and Tian He.ESC:Energy Synchronized Communication in Sustainable Sensor Networks, in ICNP ' 09) by introducing One transparent middleware at utmost reduces Network Transmission Delays, thus extends node lifetime.
For many sensor networks, under low duty ratio real-time Data Transmission be affect of node energy crucial because of Element.Scholars more both domestic and external propose some solutions also for the problem such as residue energy of node and link-quality. PTW(Xue Yang and Nitin H.Vaidya.A Wakeup Scheme for Sensor Networks:Achieving Balance Between Energy Saving and End-to-end Delay, in RTAS ' 04) describe one at energy Consumption saves the balance method between end-to-end delay.It addition, prior art also proposed SDC algorithm (Y.Gu, T.He, M.Lin,and J.Xu,“Spatiotemporal Delay Control for Low-Duty-Cycle Sensor Networks, " in RTSS ' 09,2009), this algorithm attempts to use a kind of general method to reduce the delay of network.SDC calculates Method enables to delay and controls in the case of certain condition, saves the energy expenditure of node.But this algorithm remains In place of some shortcomings, as do not considered, the data that link-quality may cause retransmit, thus cause internodal transmission delay Increase.
The research in the past increased extras, not accounting for link-quality is different, and the present invention proposes one in a WSN Node dormancy dispatching method based on link-quality, its core concept is on the basis of considering link-quality, with minimum energy Consumption cost meets the delay requirement of communication.
Summary of the invention
The technical problem to be solved is to provide a kind of node dormancy dispatching method based on link-quality, is used for Prior art extras are many, do not account for the problems such as link-quality in solution.
The technical scheme is that a kind of node dormancy dispatching party based on link-quality Method, including:
Step 1, detects the link-quality between each adjacent node on link;
Step 2, in conjunction with the link-quality detected, calculates the source node delay desired values to destination node, if this delay Expected value more than the delay threshold value of application request, then performs step 3, and otherwise flow process terminates;
Step 3, from the beginning of increasing time slot of once reviving, calculates the source node after increasing time slot of reviving successively and saves to purpose The minimum delay expected value of point, until the minimum delay expected value calculated is less than or equal to postponing threshold value, or until increases Soviet Union The number of times of awake time slot is equal to the jumping figure needed for source node to destination node.
On the basis of technique scheme, the present invention can also do following improvement.
Further, in described step 2, calculate the source node delay to destination node according to initial node scheduling table Expected value, and described node scheduling table includes time started and the time slot size of reviving of node that node revives.
Further, the time slot size of reviving of node is 1 second.
Further, calculating source node in described step 2 to the detailed process of the delay desired values of destination node is: assume source Node is i, destination node be j, now source node i and destination node j be adjacent node, useRepresent adjacent node i and Increase the delay desired values after time slot of once reviving between j, then have:
In formula, n is that data retransmit number, NmaxFor maximum retransmission,It is right to representTake downwards It is whole,Representing and round n/2 downwards, T is node i and the working cycle of node j, pijSave for source node i to purpose Link-quality between some j,Represent the recovery time of i,Represent the recovery time of j.
Further, described step 3 calculates the minimum delay to destination node of the source node after increasing time slot of reviving successively The method of expected value specifically includes: assuming that source node is i, destination node is j, from the beginning of increasing time slot of once reviving, if increasing h The source node i that secondary time slot of reviving can reach to the minimum delay expected value of destination node j isWherein m represents source Node i is less than or equal to m to the jumping figure required for destination node j, the value of h, then has:
E [ D i j m , h ( t ) ] = m i n E [ D i j m , h ( t ) ] , E [ d i 1 ( t ) ] ⊕ E [ D 1 j m - 1 , h ( t ′ ) ] , E [ d i 1 1 ( t ) ] ⊕ E [ D 1 j m - 1 , h - 1 ( t ′ ) ] h > 0 , m > 0 ,
In formula,Represent the newly-increased delay desired values between time slot posterior nodal point i and j of reviving;Be a combinatorial formula, represent and do not increase, at source node, time slot of reviving, and from source node to In the case of node j adds time slot number of times of reviving h time, node i is to the delay desired values of node j, and t' represents that node j receives number Moment according to bag;It is a combinatorial formula, represents that source node adds time slot of once reviving, And the delay desired values of the time slot posterior nodal point i to node j that revives is added h-1 time from source node to node j, t' represents node j-1 Receive the moment of packet.
The invention has the beneficial effects as follows: the present invention is on the basis of considering link-quality so that node is by inserting time slot Reduce end-to-end between delay, meet system restriction postpone.Further, the present invention can be to limit the node dormancy under postponing Dispatching algorithm provides certain Research foundation, provides certain simultaneously and think in terms of extended network lifecycle by Energy-aware Road.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of node dormancy dispatching method based on link-quality of the present invention;
Fig. 2 is the network mode schematic diagram not considering link-quality in the embodiment of the present invention;
Fig. 3 is the network mode schematic diagram considering link-quality in the embodiment of the present invention;
Fig. 4 is that the embodiment of the present invention uses the exemplary plot increasing time slot reduction delay of reviving.
Detailed description of the invention
Being described principle and the feature of the present invention below in conjunction with accompanying drawing, example is served only for explaining the present invention, and Non-for limiting the scope of the present invention.
As it is shown in figure 1, this gives a kind of node dormancy dispatching method based on link-quality, including:
Step 1, detects the link-quality between each adjacent node on link;
Step 2, in conjunction with the link-quality detected, calculates the source node delay desired values to destination node, if this delay Expected value more than the delay threshold value of application request, then performs step 3, and otherwise flow process terminates;
Step 3, from the beginning of increasing time slot of once reviving, calculates the source node after increasing time slot of reviving successively and saves to purpose The minimum delay expected value of point, until the minimum delay expected value calculated is less than or equal to postponing threshold value, or until increases Soviet Union The number of times of awake time slot is equal to the jumping figure needed for source node to destination node.
Based on above-mentioned steps, the specific implementation process of the present embodiment is as described below.
Assuming that the time slot of reviving each time of node i can use two tuplesRepresent, whereinRepresent the jth time Soviet Union of node i Wake up,Represent the persistent period that node i jth time is revived.Assuming that in T time, node is revived M time altogether, then the work of node Timetable is expressed asDescribe for simplifying, it is assumed that node cycling, each is revived The size of time slot is τ (τ represents minimum time of day), and the working cycle of node is T, then the operating schedule of node i can It is expressed as Γi={ (ti, 1) }, wherein tiThe time started that expression node is revived, such as: when the working cycle of node is 100, { (1,1) } represent node will 1,101,201 ... revive 1 τ respectively.Fig. 2 is the network transmission mould not considering link-quality Type, the working cycle of node A can use ΓA={ (1,1) } represents, in like manner ΓB={ (5,1) }, ΓC={ (8,1) }, false here The value determining τ is 1 second.In the case of not considering link-quality, the transmission delay of node A to node C is 7s.
In actual application, the data transmission success between true environment link is not 100%, and Fig. 3 is consideration The situation of link-quality, it is assumed that the link-quality between A, B is 0.8, the link-quality between B, C is 0.6.For the side of description Just, it is assumed hereinafter that node slot time is all 1 second, then the scheduling table of node i is reduced to
Because it is possible that the situation of data unsuccessful transmission, needing to introduce data and retransmit.
If source node S is followed successively by the node of process on a paths between destination node D1 should The dormancy delay desired values that paths carries out data transmission from source node to destination node at moment t is:
E [ D S D ( t ) ] = Σ n s 1 = 0 N max - 1 Σ n 12 = 0 N max - 1 ... Σ n ( m - 2 ) ( m - 1 ) = 0 N max - 1 Σ n ( m - 1 ) D = 0 N max - 1 p s 1 ( 1 - p s 1 ) n s 1 p 12 ( 1 - p 12 ) n 12 ... p ( m - 2 ) ( m - 1 ) ( 1 - p ( m - 2 ) ( m - 1 ) ) n ( m - 2 ) ( m - 1 ) p ( m - 1 ) j ( 1 - p ( n - 1 ) D ) n ( m - 1 ) D × [ d s 1 ( t ) + d 12 ( t 1 ) + ... + d ( m - 2 ) ( m - 1 ) ( t m - 2 ) + d ( m - 1 ) D ( t m - 1 ) + T × ( n s 1 + n 12 + ... + n ( m - 2 ) ( m - 1 ) + n ( m - 1 ) D ) ] - - - ( 1 )
In formula, NmaxRepresent maximum retransmission, ps1、p12…p(m-2)(m-1)、p(m-1)jRepresent corresponding link-quality, ds1 (t)、d12(t1)…d(m-2)(m-1)(tm-2)、d(m-1)D(tm-1) representing the transmission delay between each adjacent node, T represents the cycle.
As shown in Figure 3, it is assumed that cycle T=100s, maximum retransmission is Nmax=3, by calculate can draw A, C it Between transmission delay be E [DAC(1)]=70.32S.According to formula (1), the process that can calculate is:
E [ D A C ( 1 ) ] = Σ n A B = 0 2 Σ n B C = 0 2 0.8 × ( 1 - 0.8 ) n A B × 0.6 × ( 1 - 0.6 ) n B C × ( 4 + 100 · n A B + 3 + 100 · n B C ) = 0.8 × 0.6 × ( 4 + 3 ) + 0.8 × 0.6 × 0.4 × ( 4 + 3 + 100 ) + 0.8 × 0.6 × 0.4 2 × ( 4 + 3 + 200 ) + 0.8 × 0.2 × 0.6 × ( 4 + 100 + 3 ) + 0.8 × 0.2 × 0.6 × 0.4 × ( 4 + 100 + 3 + 100 ) + 0.8 × 0.2 × 0.6 × 0.4 2 × ( 4 + 100 + 3 + 200 ) + 0.8 × 0.2 2 × 0.6 × ( 4 + 200 + 3 ) + 0.8 × 0.2 2 × 0.6 × 0.4 × ( 4 + 200 + 3 + 100 ) + 0.8 × 0.2 2 × 0.6 × 0.4 2 × ( 4 + 200 + 3 + 200 ) ≈ 70.32 ( s )
It is apparent from, it is considered to the expectation transmission time after link-quality is more than original 10 times.Above-mentioned result of calculation is base In double bounce path computing, along with the increase of jumping figure between end-to-end in network, dormancy postpones also will become much larger.Thus may be used Seeing, link in network quality is huge on the impact of end-to-end delay.
Owing to the insecure WSN of link-quality postponing the requirement being unable to greatly meet reality application, can be by increasing Node time slot of reviving is to reduce delay, and Fig. 4 is and increases time slot of reviving and reduce the exemplary plot postponed.Through calculating, It is about 69.02s, node B at initial condition lower node A to the dormancy delay desired values that node B transmits and increases time slot of once reviving After this value become 11.55s, postpone to decrease about 83%.
Typically, for neighbor node i and j, once revive time slot to reduce dormancy delay, i.e. when node j needs to increase The working cycle of node j byBecomeTime.IfSo can draw newly-increased Revive the delay desired values between time slot posterior nodal point i and jFor formula:
In formula, n is that data retransmit number, NmaxFor maximum retransmission,It is right to representTake downwards It is whole,Representing and round n/2 downwards, T is node i and the working cycle of node j, pijSave for source node i to purpose Link-quality between some j,Represent the recovery time of i,Represent the recovery time of j.
Postpone to node j dormancy end to end to calculate node i, useRepresent when node i is received in t To packet and be ready for sending to node j, the minimum dormancy period of delay increasing on this paths that time slot of reviving for h time can reach Prestige value.M is that on this paths, node i is less than or equal to m to the jumping figure required for node j, the value of h.This algorithm is divided into following two Step:
The first step, according to initial node scheduling table, utilizes formula (2) to calculate from source node i to destination node j Delay desired valuesIf this expected value is less than or equal to delay threshold value B of application request, i.e.Then think and met the requirement of actual application, it is not necessary to the extra wake up operations increasing node, this algorithm Exit.Otherwise enter second step.
Second step, from the beginning of increasing by 1 time slot of reviving, calculates successively and increases the minimum that time slot of reviving for h time can reach Delay desired valuesUntil meeting inequalityOr h=m, at this moment the value of h is minimum, institute Energy expenditure to be increased also is minimum.This ensures that there and postpone (B) satisfied restriction from source node to destination node In the case of, the energy consumed is minimum.
In second step, for neighbor node i and j, if not increasing the wake up operations of node j, then node i and node j Postpone as original value.In a working cycle, if the moment of reviving of node j is not the moment of reviving big 1 than node i If individual time slot, then the dormancy that the time slot number of times of reviving increasing by a minor node j will reduce between node i and node j postpones. Therefore can show that the delay desired values of two adjacent nodes meets below equation:
E [ D i j m , h ( t ) ] = E [ d i j ( t ) ] m = 1 , h = 0 E [ d i j 1 ( t ) ] m = 1 , h = 1 - - - ( 3 )
In formula, E [dij(t)] can be drawn by the algorithm of formula (1),Can be drawn by the algorithm of formula (2).
Considering now general situation, node i and node j are not neighbor node, and node i to send data to node j Information has to pass through other m-1 hop node on certain paths.So for node j, finally send over from source node i Data include following two situation to node j:
First, during packet is from source node to destination node, node 1 does not increase time slot of reviving, and is to maintain Working method originally is constant, is then altogether adding h time slot of reviving from node 1 to node j;
Second, packet is from source node to its next-hop node 1, and in order to reduce dormancy delay desired values, node 1 increases by one Secondary time slot of reviving, adds, to node j, time slot of reviving from node 1 h-1 time.
According to the two kinds of situations being considered above and formula (2), after the time slot that can go out to increase to revive for h time by iterative computation, The expected value that end-to-end dormancy between node i and node j postpones.Therefore, it can to obtain below equation obtain increasing and revive for h time The source node i that time slot can reach to the minimum delay expected value of destination node j is
E [ D i j m , h ( t ) ] = m i n E [ D i j m , h ( t ) ] , E [ d i 1 ( t ) ] ⊕ E [ D 1 j m - 1 , h ( t ′ ) ] , E [ d i 1 1 ( t ) ] ⊕ E [ D 1 j m - 1 , h - 1 ( t ′ ) ] h > 0 , m > 0 - - - ( 4 )
In formula,Represent the newly-increased delay desired values between time slot posterior nodal point i and j of reviving;Be a combinatorial formula, represent and do not increase, at source node, time slot of reviving, and from source node to In the case of node j adds time slot number of times of reviving h time, node i is to the delay desired values of node j, and t' represents that node j receives number Moment according to bag;Being a combinatorial formula, expression source node adds when once reviving Gap, and the delay desired values of the time slot posterior nodal point i to node j that revives is added h-1 time from source node to node j, t' represents node J-1 receives the moment of packet,Occurrence acquisition can in conjunction with formula (1) to formula (3) obtain.
The foregoing is only presently preferred embodiments of the present invention, not in order to limit the present invention, all spirit in the present invention and Within principle, any modification, equivalent substitution and improvement etc. made, should be included within the scope of the present invention.

Claims (4)

1. a node dormancy dispatching method based on link-quality, it is characterised in that including:
Step 1, detects the link-quality between each adjacent node on link;
Step 2, in conjunction with the link-quality detected, according to link transmission number of times, the period of delay of calculating source node to destination node Prestige value, if this delay desired values is more than the delay threshold value of application request, then performs step 3, and otherwise flow process terminates;
Step 3, from the beginning of increasing time slot of once reviving, according to link transmission number of times, calculates after increasing time slot of reviving successively Source node is to the minimum delay expected value of destination node, until the minimum delay expected value calculated is less than or equal to postponing threshold Value, or until increase the number of times of time slot of reviving equal to the jumping figure needed for source node to destination node.
Node dormancy dispatching method the most according to claim 1, it is characterised in that in described step 2, according to initial joint Point scheduling table calculates the source node delay desired values to destination node, and described node scheduling table includes that node is revived Time started and the time slot size of reviving of node.
Node dormancy dispatching method the most according to claim 2, it is characterised in that the time slot size of reviving of node is 1 second.
Node dormancy dispatching method the most according to claim 1, it is characterised in that calculate source node in described step 2 and arrive The detailed process of the delay desired values of destination node is: assuming that source node is i, destination node is j, now source node i and purpose Node j is adjacent node, usesRepresent the delay desired values after increasing time slot of once reviving between adjacent node i and j, Then have:
In formula, n is that data retransmit number, NmaxFor maximum retransmission,It is right to representRound downwards,Representing and round n/2 downwards, T is node i and the working cycle of node j, pijFor source node i to destination node j Between link-quality,Represent the recovery time of i,Represent the recovery time of node j.
CN201310253355.7A 2013-06-24 2013-06-24 Node dormancy scheduling method based on link quality Expired - Fee Related CN103327518B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310253355.7A CN103327518B (en) 2013-06-24 2013-06-24 Node dormancy scheduling method based on link quality

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310253355.7A CN103327518B (en) 2013-06-24 2013-06-24 Node dormancy scheduling method based on link quality

Publications (2)

Publication Number Publication Date
CN103327518A CN103327518A (en) 2013-09-25
CN103327518B true CN103327518B (en) 2017-01-11

Family

ID=49195996

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310253355.7A Expired - Fee Related CN103327518B (en) 2013-06-24 2013-06-24 Node dormancy scheduling method based on link quality

Country Status (1)

Country Link
CN (1) CN103327518B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104703259B (en) * 2013-12-10 2018-12-14 深圳先进技术研究院 Low power consuming sensor node sleep method and system
CN108337711B (en) * 2018-03-23 2021-08-20 浙江树人学院 Mobile sensing data routing method of mobile wireless sensor network
CN111758231A (en) * 2019-06-25 2020-10-09 深圳市大疆创新科技有限公司 Data self-adaptive retransmission method, remote control device, aircraft and movable platform

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103024859A (en) * 2012-11-30 2013-04-03 无锡清华信息科学与技术国家实验室物联网技术中心 Data collection method of minimum overhead in low duty-cycle wireless sensor network

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8542610B2 (en) * 2011-03-31 2013-09-24 Motorola Solutions, Inc. Adaptive power saving for a mobile subscriber unit in a wireless communication network

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103024859A (en) * 2012-11-30 2013-04-03 无锡清华信息科学与技术国家实验室物联网技术中心 Data collection method of minimum overhead in low duty-cycle wireless sensor network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
《无线传感器网络中的调度算法研究》;康波 等;《计算机科学》;20080229;第35卷(第2期);全文 *

Also Published As

Publication number Publication date
CN103327518A (en) 2013-09-25

Similar Documents

Publication Publication Date Title
Yang et al. Wireless body area networks MAC protocol for energy efficiency and extending lifetime
CN101895956B (en) Data transmission method of multilayer distributed wireless sensor network
Altman et al. Optimal activation and transmission control in delay tolerant networks
CN107682905B (en) A kind of stellar chain type wireless sensor network is without layer communication method
CN103906207B (en) Wireless sensor network data transmission method based on self adaptation awakening technology on demand
CN102868602A (en) Improved routing method based on social correlation degree in delay tolerant network
CN102651898A (en) Delay-constrained and energy-efficient online routing for asynchronous sensor networks
CN106131918A (en) The associating Path selection of energy acquisition node and power distribution method in wireless sense network
CN104469878B (en) A method of the clustering routing suitable for contactless charging WSN network
CN106060871A (en) Wireless networking and data forwarding method based on low power consumption and micro power
CN104093188A (en) Novel cluster head election method for wireless sensor networks
CN102740365A (en) Single-stream bulk data acquisition method suitable for wireless sensor network
CN103327518B (en) Node dormancy scheduling method based on link quality
CN107222903A (en) Method of data capture based on constructive interference in asynchronous low duty ratio WSN
CN113453304A (en) Self-adaptive synchronous MAC protocol based on virtual clustering and path reservation
CN102946298B (en) Based on the data repeating method of opportunistic network coding in wireless sensor network
CN103298038A (en) IEEE 802.11 WLAN (wireless local area network) low-delay energy-saving method
CN103024859B (en) The method of data capture of minimal time delay in low duty ratio wireless sensor network
Lu et al. Opportunistic forwarding in energy harvesting mobile delay tolerant networks
Deng et al. Early wake-up ahead node for fast code dissemination in wireless sensor networks
CN105208671B (en) The implementation method of high channel utilization rate H-MAC agreements for wireless sensor network
CN104010339B (en) A kind of WSN cluster-dividing methods based on packet loss rate measurement
CN106714284A (en) Sequential sleep and wakeup method for wireless chain type topological network
CN102917467B (en) Asynchronous reservation channel access method of wireless sensor network
Park et al. Design and analysis of asynchronous wakeup for wireless sensor networks

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20170111

Termination date: 20190624

CF01 Termination of patent right due to non-payment of annual fee